Package 'jinx'

Description

Caches the allowed-bases list in KV (same AIRTABLE_BASES namespace and 1h TTL the Worker used) rather than calling Airtable's Meta API on every webhook. R port of airtable_base_is_allowed()/ airtable_allowed_bases_get() from the deleted worker/src/airtable-meta.js.

Usage

airtable_base_allowed(
  base_id,
  namespace_id = airtable_bases_namespace_id(),
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  api_token = Sys.getenv("CLOUDFLARE_API_TOKEN"),
  api_key = Sys.getenv("AIRTABLE_API_KEY")
)

Arguments

Value

TRUE if the base is within the token's scope.

Description

R's first Airtable write primitive - everything in R/airtable-sync.R is read-only. R port of airtable_record_update() from the deleted worker/src/airtable-invite.js.

Usage

airtable_record_update(
  base_id,
  table_id,
  record_id,
  fields,
  api_key = Sys.getenv("AIRTABLE_API_KEY")
)

Arguments

Value

Invisibly, TRUE.

Description

The airtable_webhook event handler registered in jinx_events(). Checks the base is within the configured Airtable token's scope, builds the approval-request card, and posts it to the community-invite channel. R port of the remainder of airtable_webhook_handle() (after shared-secret verification, which stays in the Worker) from the deleted worker/src/airtable-invite.js.

Usage

airtable_webhook_process(
  email,
  name = NULL,
  chapter = NULL,
  record_id,
  base_id,
  table_id,
  channel = Sys.getenv("SLACK_COMMUNITY_INVITE_CHANNEL")
)

Arguments

Value

Invisibly, TRUE if posted, FALSE if the base was rejected.

Description

Queries the GitHub API for repository activity metrics across all non-meetup repos in the organization.

Usage

analytics_collect_chapter_activity(
  org = "rladies",
  months = 12,
  exclude_pattern = "^meetup-"
)

Arguments

Value

Data frame with columns: chapter, month, commits, prs, issues.

Description

Collect contributor growth data

Usage

analytics_collect_contributor_growth(org = "rladies", months = 12)

Arguments

Value

Data frame with columns: month, new_contributors, total_contributors, active_repos.

Description

Calculates month-over-month changes and rolling averages from chapter activity data.

Usage

analytics_compute_trends(activity_data)

Arguments

Value

Data frame with trend columns added.

Description

Format analytics as markdown

Usage

analytics_format_markdown(trends, contributor_growth)

Arguments

Value

Formatted markdown string.

Description

Orchestrates data collection, trend computation, and formatting.

Usage

analytics_generate_dashboard(org = "rladies", months = 12, output_path = NULL)

Arguments

Value

Named list with trends, growth, and markdown (invisibly).

Description

Create a formatted announcement message

Usage

announce_create_message(frontmatter, uri)

Arguments

Value

Formatted message string.

Description

Reads YAML front matter from a blog post, creates an announcement message, and posts to selected platforms.

Usage

announce_post(
  post,
  platforms = c("bluesky", "linkedin", "mastodon"),
  newsletter = TRUE
)

Arguments

Value

Invisibly returns a named list with the posting result for each platform.

Description

Post an announcement to Bluesky

Usage

announce_post_bluesky(text, image = NULL, image_alt = NULL)

Arguments

Value

Post response (invisibly).

Description

Post an announcement to Mastodon

Usage

announce_post_mastodon(text, image = NULL, image_alt = NULL)

Arguments

Value

Post response (invisibly).

Description

Posts as the authenticated user (jinx[bot] when using a GitHub App token).

Usage

announce_post_reply(owner, repo, issue_number, body)

Arguments

Value

The API response (invisibly).

Description

Send a newsletter via ConvertKit

Usage

announce_send_newsletter(frontmatter, url)

Arguments

Value

API response (invisibly).

Description

Builds an entry from the URL's metadata and opens a PR adding it to the awesome-rladies-creations content directory. If an entry for the URL's domain already exists, no PR is opened.

Usage

blog_add_pr(
  url,
  org = "rladies",
  repo = "awesome-rladies-creations",
  content_path = "data/content",
  language = "en",
  author_name = NULL,
  base = "main"
)

Arguments

Value

A list with status ("created" or "exists"), filename, and url (the PR URL when created).

Description

Check blog URLs and RSS feeds for broken links

Usage

blog_check_links(blogs_path)

Arguments

Value

A data frame with columns file, url, rss_feed, url_status, rss_status.

Description

Lists the blog entry JSON files in the content directory via the GitHub API and checks each entry's url and rss_feed for broken links.

Usage

blog_check_links_repo(
  org = "rladies",
  repo = "awesome-rladies-creations",
  content_path = "data/content"
)

Arguments

Value

A data frame with columns file, url, rss_feed, url_status, rss_status.

Description

Fetches OpenGraph metadata from the URL and writes a JSON entry compatible with awesome-rladies-creations format.

Usage

blog_create_entry(url, language = "en", author_name, output_dir = ".")

Arguments

Value

File path of the created JSON (invisibly).

Description

The heading, a linked bold title and the post's own description go in one section; the preview image follows as its own block so Slack renders it full width rather than as a thumbnail; the byline sits in a context block, which Slack renders small and grey so it reads as attribution rather than as part of the post.

Usage

blog_feed_blocks(item, source, og = list())

Arguments

Details

Every piece of text here comes from a feed or a page that contributors control, so all of it is escaped - the link markup around the title is the only markup the message is allowed to introduce.

Value

A list of Slack Block Kit blocks.

Description

Polls each source in turn, keeping going when an individual feed is down so one dead blog can't silence the rest, then returns the items sorted oldest first so a channel reads in publication order.

Usage

blog_feed_collect(
  entries = NULL,
  max_age_days = 14,
  now = as.numeric(Sys.time())
)

Arguments

Details

Deliberately knows nothing about what any workspace has already announced: both blog channels want the same posts, so a run polls the feeds once and each workspace filters the result against its own seen-set with blog_feed_pending(). Polling per workspace would double the requests every contributor's blog receives.

Value

A list with posts (a list of item/source pairs, oldest first), feedless (titles with no rss_feed), and sources (how many feeds were polled).

Description

Titles and author names come from feeds and curated JSON that contributors control, so both are escaped before they go into a broadcast message - that neutralises injected links and ⁠<!channel>⁠/⁠<!everyone>⁠ mass-pings while leaving the message's own link markup intact.

Usage

blog_feed_format(item, source)

Arguments

Value

Character scalar Slack mrkdwn message.

Description

Matched on local-name() rather than a namespace prefix: the feeds in the curated list are a mix of RSS 2.0, Atom, and RSS 1.0/RDF written by half a dozen site generators, and only the element names are reliably shared between them.

Usage

blog_feed_items(xml, base = NULL, max_items = 25L)

Arguments

Value

A data frame with columns id, title, link, and date (a unix timestamp, 0 when the feed gave no usable date).

Description

Drops anything already announced and anything older than max_age_days

• a blog joining the curated list arrives with its whole back catalogue in the feed, and without the age gate its first run would announce years of posts at once. Items with no usable date are treated as too old for the same reason.

Usage

blog_feed_new_items(
  items,
  seen = character(),
  max_age_days = 14,
  now = as.numeric(Sys.time())
)

Arguments

Value

A subset of items, oldest first.

Description

One extra request per new post - not per feed, and not per item - so at a post a week this is nothing next to polling the feeds themselves. Every failure returns empty: a post with no preview is still worth announcing, so nothing here is allowed to abort the run.

Usage

blog_feed_og(url)

Arguments

Value

A list with description, image, and image_alt, each a character scalar that may be "".

Description

Keep the posts one workspace still owes its channel

Usage

blog_feed_pending(posts, seen = character(), limit = 20L)

Arguments

Value

The subset still to announce, oldest first.

Description

Replaces the Slack Feed app's per-blog subscriptions with one run driven by awesome-rladies-creations: a blog added to the curated list is announced without anyone touching Slack.

Usage

blog_feed_post(
  workspace = c("community", "organiser"),
  channel = env_default("SLACK_BLOG_CHANNEL", "blogs-by-rladies"),
  dry_run = FALSE,
  posts = NULL,
  entries = NULL,
  max_age_days = 14,
  limit = 20L,
  slack_token = NULL,
  og_fetch = blog_feed_og_memo(),
  namespace_id = slack_tokens_namespace_id(),
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  api_token = Sys.getenv("CLOUDFLARE_API_TOKEN")
)

Arguments

Details

Announced ids are recorded in KV per workspace, and only after a successful post - a failed run re-announces nothing and loses nothing.

Use blog_feed_run() to serve both workspaces, which polls the feeds once for the pair rather than once each.

Value

Invisibly, the number of posts announced.

Description

The scheduled entry point. Both channels want the same posts, so the feeds are polled once and each workspace is then filtered against its own seen-set - at a ten-minute cadence, polling per workspace would double the requests every contributor's blog receives for no gain.

Usage

blog_feed_run(
  workspaces = c("community", "organiser"),
  channel = env_default("SLACK_BLOG_CHANNEL", "blogs-by-rladies"),
  dry_run = FALSE,
  seed = FALSE,
  entries = NULL,
  max_age_days = 14,
  limit = 20L,
  namespace_id = slack_tokens_namespace_id(),
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  api_token = Sys.getenv("CLOUDFLARE_API_TOKEN")
)

Arguments

Details

One workspace failing does not stop the others: each is attempted, and the error is re-raised at the end so the run still goes red.

Value

Invisibly, a named integer of posts announced per workspace.

Description

An entry is pollable when it carries an rss_feed, whatever its type

• YouTube channels and plain websites serve the same Atom/RSS XML as blogs do. Entries without a feed can't be polled at all, so they are returned separately for the run to report rather than dropped silently.

Usage

blog_feed_sources(entries)

Arguments

Value

A list with sources (a data frame with columns title, site, feed, author, and type) and feedless (a character vector of titles).

Description

Format Workers invocation metrics as a markdown table

Usage

cf_ops_format_workers_report(df)

Arguments

Value

Character string with markdown-formatted report.

Description

Read-only, for interactive/console use during an incident — see cf_ops_list_kv_keys() for why this isn't wired into a ⁠/jinx⁠ command.

Usage

cf_ops_get_kv_value(
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  namespace_id,
  key_name,
  token = cf_ops_token()
)

Arguments

Value

Character string with the stored value.

Description

Wraps cloudflarer::cf_delete_kv_value(). See cf_ops_kv_put() for why this is narrowly scoped rather than a general-purpose tool.

Usage

cf_ops_kv_delete(
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  namespace_id,
  key_name,
  token = cf_ops_token()
)

Arguments

Value

The API response (invisibly).

Description

Wraps cloudflarer::cf_put_kv_value(). Unlike cf_ops_list_kv_keys()/ cf_ops_get_kv_value(), this is not a general-purpose "read/write any key" tool exposed to a ⁠/jinx⁠ command — it's called internally by event/command handlers against a small set of hardcoded key patterns (e.g. pending_link:{email}, channel_index:{team_id}), the same scoping the exfiltration concern in those two functions' docs warns about.

Usage

cf_ops_kv_put(
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  namespace_id,
  key_name,
  value,
  ttl_seconds = NULL,
  token = cf_ops_token()
)

Arguments

Value

The API response (invisibly).

Description

Read-only, for interactive/console use during an incident — not wired into a ⁠/jinx⁠ command, since KV namespaces on this account (SLACK_TOKENS, AIRTABLE_BASES) hold operational secrets, and a chat-triggered arbitrary-namespace read is a real exfiltration path.

Usage

cf_ops_list_kv_keys(
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  namespace_id,
  prefix = NULL,
  token = cf_ops_token()
)

Arguments

Value

Data frame of key records.

Description

Thin wrapper over cloudflarer::cf_list_workers_scripts().

Usage

cf_ops_list_workers_scripts(
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  token = cf_ops_token()
)

Arguments

Value

Data frame of Workers script records.

Description

Thin wrapper over cloudflarer::cf_purge_cache() that deliberately does not expose purge_everything — a maintainer who genuinely needs a full-zone wipe calls cloudflarer::cf_purge_cache(purge_everything = TRUE) directly. That extra step is intentional friction on the most dangerous call this integration can make.

Usage

cf_ops_purge_cache(
  zone_id = Sys.getenv("CLOUDFLARE_ZONE_ID"),
  files = NULL,
  prefixes = NULL,
  token = cf_ops_token()
)

Arguments

Value

A named list with the purge job id.

Description

Thin wrapper over cloudflarer::cf_workers_invocations(), defaulting script_name to "jinx" (matches wrangler.jsonc's worker name) rather than requiring the caller to know it.

Usage

cf_ops_workers_invocations(
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  since = Sys.Date() - 1,
  until = Sys.Date(),
  script_name = "jinx",
  token = cf_ops_token()
)

Arguments

Value

Data frame with columns date, script, requests, errors, subrequests, cpu_p50_us, cpu_p99_us.

Description

Check CFP deadlines and post reminders

Usage

cfp_check_deadlines(org = "rladies", repo = "global-team", warn_days = 7)

Arguments

Value

Data frame of approaching CFPs (invisibly).

Description

Create a CFP tracking issue

Usage

cfp_create_issue(
  conference,
  deadline,
  url,
  topics = character(0),
  org = "rladies",
  repo = "global-team"
)

Arguments

Value

Issue URL (invisibly).

Description

List open CFPs tracked as GitHub issues

Usage

cfp_list_open(org = "rladies", repo = "global-team")

Arguments

Value

Data frame with columns: conference, deadline, url, number, status.

Description

Adds a speaker recommendation as a comment on the CFP tracking issue.

Usage

cfp_recommend_speaker(
  conference,
  speaker_name,
  speaker_github = NULL,
  expertise = character(0),
  org = "rladies",
  repo = "global-team"
)

Arguments

Value

Comment URL (invisibly).

Description

Only rows with status "apply" are sent to Slack. Everything else is returned untouched so the caller keeps a full record of what was skipped and why.

Usage

channel_copy_apply(
  plan,
  token,
  dry_run = TRUE,
  skip = character(),
  join = TRUE,
  include_drift = FALSE,
  user_token = NULL
)

Arguments

Value

plan with applied and error columns added.

Description

Compares the reviewed proposals against the workspace's live state and classifies every intended change. A row is "unchanged" when Slack already holds the proposed value, "drift" when the live value no longer matches what the review recorded (someone edited it since, so the proposal may be stale), "missing" when the channel is not in the workspace, and "apply" otherwise.

Usage

channel_copy_plan(
  workspace = c("organiser", "community"),
  index = NULL,
  proposals = NULL,
  renames = NULL
)

Arguments

Details

Renames are planned last so a rename cannot hide the channel from the copy rows in the same pass.

Value

A data frame of planned changes.

Description

Read the reviewed channel copy proposals

Usage

channel_copy_proposals(workspace = c("organiser", "community"), path = NULL)

Arguments

Value

A data frame of proposals for that workspace, excluding channels that do not exist yet.

Description

Callers resolving several channel names at once (e.g. welcome_message_render()'s starter-channel list) should call this once and look names up in the result, rather than calling slack_channel_id_lookup() per name - each call independently checks the KV cache, so looking up N names one-by-one means N redundant KV reads (and N cache-miss races) for data that doesn't change between them.

Usage

channel_index_load(
  team_id,
  workspace,
  namespace_id = slack_tokens_namespace_id(),
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  api_token = Sys.getenv("CLOUDFLARE_API_TOKEN")
)

Arguments

Value

A named list mapping channel name to channel id, or NULL if the channel list couldn't be fetched.

Description

Lists the community workspace's public channels, keeps those with a description, picks the next one round-robin (state stored in KV), and drafts a Jinx-voiced invitation for it - falling back to the plain description if the model call fails.

Usage

channel_promo_build(
  team_id = Sys.getenv("SLACK_COMMUNITY_TEAM_ID"),
  target_channel = env_default("SLACK_PROMO_CHANNEL", "general"),
  skip = promo_skip_channels(),
  namespace_id = slack_tokens_namespace_id(),
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  api_token = Sys.getenv("CLOUDFLARE_API_TOKEN"),
  model = workers_ai_chat_model()
)

Arguments

Value

A list with text (the Slack message), channel_id, channel_name, and recent (the updated round-robin state to persist), or NULL when no channel is eligible.

Description

Links the featured channel in the header and appends the invitation. The blurb is escaped with escape_markdown() before it goes out: a channel's description is user-controlled and flows through the model into a broadcast message, so this neutralises injected links and ⁠<!channel>⁠/⁠<!everyone>⁠ mass-pings while leaving ⁠*bold*⁠/⁠_italic_⁠ intact.

Usage

channel_promo_format(id, name, blurb)

Arguments

Value

Character scalar Slack mrkdwn message.

Description

Posts one channel's spotlight to the target channel, then records the pick in KV so next week's run features a different channel. The round-robin state is only advanced after a successful post, so a failed run doesn't burn a channel's turn.

Usage

channel_promo_post(
  team_id = Sys.getenv("SLACK_COMMUNITY_TEAM_ID"),
  target_channel = env_default("SLACK_PROMO_CHANNEL", "general"),
  slack_token = slack_bot_token("community"),
  skip = promo_skip_channels(),
  namespace_id = slack_tokens_namespace_id(),
  account_id = Sys.getenv("CLOUDFLARE_ACCOUNT_ID"),
  api_token = Sys.getenv("CLOUDFLARE_API_TOKEN"),
  model = workers_ai_chat_model()
)

Arguments

Value

Invisibly, TRUE if a spotlight was posted, FALSE if there was nothing eligible to promote.

Description

Read the reviewed channel rename proposals

Usage

channel_rename_proposals(workspace = c("organiser", "community"), path = NULL)

Arguments

Value

A data frame of old/new channel names for that workspace.

Description

Identifies chapters that have been inactive (no events) for a specified number of months.

Usage

chapter_check_health(
  months = 12,
  org = "rladies",
  data_repo = "meetup_archive"
)

Arguments

Value

A data frame with columns chapter, last_event, status, and months_inactive.

Description

Tick a checklist item on an onboarding issue

Usage

chapter_checklist_tick(
  issue_number,
  pattern,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

TRUE when an item was ticked, FALSE otherwise.

Description

Sends the confirmation email from [email protected], cc'ing [email protected] so the onboarding inbox keeps the thread, and notes on the issue that it went out.

Usage

chapter_confirm_send(
  issue_number,
  to,
  city = NULL,
  email = NULL,
  meetup_url = NULL,
  force = FALSE,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Details

Refuses to send while checklist steps are outstanding, since the email tells the organisers their chapter is set up. Pass force = TRUE to override, which is occasionally right - a step may be recorded elsewhere - but it should be a decision rather than an accident.

Recipient addresses are passed in rather than read from the issue: organiser emails are deliberately not stored in the issue metadata, and this function does not put them back there either. The issue comment records how many organisers were written to, not who.

Value

The sent message id (invisibly).

Description

Generates a chapter JSON entry for the website.

Usage

chapter_create(
  city,
  country,
  organizers,
  region = NULL,
  social_media = list(),
  status = "prospective",
  output_dir = "."
)

Arguments

Value

File path of the created JSON (invisibly).

Description

Generates the chapter JSON entry, validates it against the bundled schema, and opens a PR adding it to the website. A prospective chapter has no Meetup group or chapter email yet, so both are optional and are simply left out of the entry.

Usage

chapter_create_pr(
  city,
  country,
  region = NULL,
  meetup_urlname = NULL,
  email = NULL,
  organizers = character(0),
  status = "prospective",
  social_media = list(),
  org = "rladies",
  website_repo = "rladies.github.io",
  team_reviewers = "leadership"
)

Arguments

Value

PR URL (invisibly).

Description

Opens a tracking issue in the new-chapters-onboarding repo with the full checklist for setting up a new RLadies+ chapter.

Usage

chapter_create_setup(
  city,
  country,
  organizers,
  region = NULL,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

Issue URL (invisibly).

Description

Opens a tracking issue for updating an existing chapter's infrastructure.

Usage

chapter_create_update(
  city,
  country,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

Issue URL (invisibly).

Description

Performs the first step of the chapter onboarding checklist: searches the website's data/chapters entries for a chapter in the same city, for chapters with a near-identical city name, and for chapters in the same country within radius_km of the requested city.

Usage

chapter_duplicate_check(
  city,
  country,
  region = NULL,
  radius_km = 100,
  org = "rladies",
  website_repo = "rladies.github.io"
)

Arguments

Details

Only chapters in the requested country, or whose filename contains the requested city slug, are fetched in full, so the check costs one directory listing plus a handful of file reads.

Value

A data frame with one row per match and columns city, country, status, email, meetup, match, and distance_km. match is one of "exact", "similar", or "nearby".

Description

Runs chapter_duplicate_check() and comments the rendered report on the issue. Failures are reported but never abort onboarding: the check is an aid to the team, not a gate.

Usage

chapter_duplicate_comment(
  issue_number,
  city,
  country,
  region = NULL,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

The rendered report (invisibly), or NULL on failure.

Description

Turns the output of chapter_duplicate_check() into the markdown the onboarding team reads first: whether the city already has a chapter, which nearby chapters should be cc'd, and which names are close enough to be worth a second look.

Usage

chapter_duplicate_report(matches, city, country, radius_km = 100)

Arguments

Value

Markdown body as a single string.

Description

An approval is a comment containing the approval marker, written by a confirmed member of an approving team. Bot comments never count, so jinx cannot approve its own work.

Usage

chapter_email_approvals(
  issue_number,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

Character vector of approver logins, possibly empty.

Description

Reads the chapter from the onboarding issue's own machine-readable block, checks that a member of an approving team has left an approval comment, and only then asks the worker to create the mailbox. The address is posted back on the issue and the email step of the checklist is ticked.

Usage

chapter_email_provision(
  issue_number,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Details

The approval is a separate, deliberate act on the specific issue: being allowed to run the command is not on its own enough to create an account, so a stray dispatch or a mistaken command cannot provision anything.

Value

The created address (invisibly).

Description

Matches on the issue title, which onboarding issues open as ", chapter setup".

Usage

chapter_find_issue(
  city,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

The issue number, or NULL when no open issue matches.

Description

Used as the start of the six-month window in which a chapter with no events counts as unbegun rather than inactive.

Usage

chapter_initiated_date(files, repo = ".", bulk_import = as.Date("2023-01-04"))

Arguments

Details

Two things make this fragile, and both are handled here. Renaming a chapter file would otherwise reset its date to the rename, so the lookup follows renames. A shallow clone has no history to read, so the caller's checkout needs fetch-depth: 0; without it every chapter looks as though it appeared today.

Dates before bulk_import are not meaningful: 118 chapters were added to the website in one import on 2023-01-04, so that date records the import rather than when any of those chapters started. Those come back as NA.

Value

A Date vector, NA where the date is unknown or untrusted.

Description

Asks the jinx worker to provision ⁠<city>@rladies.org⁠ in the ⁠/Chapters⁠ organisational unit. The Workspace service account key lives in the worker rather than in CI, so this call carries only the worker API key: the most a stolen CI token can do is ask for a mailbox whose name passes the worker's validation.

Usage

chapter_mailbox_create(
  city,
  given_name = NULL,
  family_name = NULL,
  base_url = "https://jinx.rladies.org",
  api_key = worker_api_key()
)

Arguments

Details

The signing and Directory API calls could be done here in R - openssl signs RS256 perfectly well - but that would require the service account key on the Actions runner that runs this function, which is the thing the current arrangement avoids. worker/src/workspace.js carries the full reasoning; it is a deliberate exception to moving worker logic into this package.

The generated password is never returned. The account is created with change-password-at-next-login set, and the onboarding team issues the handover from the Admin console.

Value

The created address (invisibly).

Description

Everything the Meetup Pro team needs in one comment: the exact group name and URL, the settings the organisational guidelines call for, and the standard description ready to paste.

Usage

chapter_meetup_brief(city, country, status = NULL)

Arguments

Value

Markdown body as a single string.

Description

The website's data/chapters entries are the source of truth for which chapters exist and which Meetup group each uses, so the sync derives its list from there rather than from a hand-maintained one that would drift the day it was written.

Usage

chapter_meetup_groups(
  org = "rladies",
  website_repo = "rladies.github.io",
  statuses = c("active", "prospective")
)

Arguments

Details

Retired chapters are excluded by default: they have no upcoming events and querying them only spends requests.

Value

A data frame with columns file, city, country, status, and urlname.

Description

Meetup uploads in two steps: a mutation registers the photo and returns a signed uploadUrl, then the bytes are sent to that URL. The mutation is named for event photos, but photoType carries a GROUP_PHOTO value and setAsMain makes it the group's image.

Usage

chapter_meetup_logo_upload(urlname, image = chapter_meetup_logo_path())

Arguments

Value

The uploaded image path reported by Meetup (invisibly).

Description

The Meetup group name for a chapter

Usage

chapter_meetup_name(city)

Arguments

Value

The group name.

Description

Reads the chapter from the issue's own machine-readable block, so in the usual case only the issue number is needed.

Usage

chapter_meetup_request(
  issue_number,
  city = NULL,
  country = NULL,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

The rendered brief (invisibly).

Description

⁠meetup.com/rladies-<city>⁠, hyphenated for multi-word cities and transliterated the way the website filenames already are. This value and the group name cannot be changed after the group is created, so getting them right first time is the whole point of rendering them here rather than having someone type them.

Usage

chapter_meetup_urlname(city)

Arguments

Value

The urlname, without the meetup.com prefix.

Description

Read the machine-readable block off an onboarding issue

Usage

chapter_meta_fetch(
  issue_number,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

A named list of chapter details, or NULL when the issue carries no block.

Description

Determines chapter status based on meetup event data. Chapters are classified as active, inactive, or unbegun based on their last event and founding date.

Usage

chapter_monitor_status(data_path = NULL, inactive_months = 6, org = "rladies")

Arguments

Value

Data frame with chapter status information.

Description

The owner of the first unchecked item, or NA when the checklist is complete or carries no owner.

Usage

chapter_onboard_blocker(state)

Arguments

Value

A team slug such as "rladies/email", or NA_character_.

Description

Parses the task list in an onboarding issue body into one row per checklist item, carrying the section it sits under and the team that section says is responsible for it. Nested items inherit their parent's owner.

Usage

chapter_onboard_state(
  issue_number,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

A data frame with columns section, owner, item, and done.

Description

Reports progress per section, names the next unchecked item and who owns it, and says plainly when everything is done.

Usage

chapter_onboard_summary(state, title = "Chapter onboarding")

Arguments

Value

Markdown body as a single string.

Description

Opens the website PR with the chapter's prospective entry, requests review from the leadership team, links the PR on the onboarding issue, and ticks the website step of the checklist.

Usage

chapter_onboard_website(
  issue_number,
  city = NULL,
  country = NULL,
  region = NULL,
  organizers = NULL,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding",
  website_repo = "rladies.github.io"
)

Arguments

Details

A prospective entry deliberately carries no chapter email and no Meetup group: neither exists at this point in onboarding, and the entry is filled in as those steps complete.

Anything left NULL is read from the issue's own machine-readable block, so in the usual case only the issue number is needed.

Value

The website PR URL (invisibly).

Description

The form's chapter field is free text and inconsistently written - "Cologne, Germany" beside "Wake Forest, North Carolina EE.UU" - so matching is by city slug appearing anywhere in it, and the caller is expected to show a human what was found rather than act on it.

Usage

chapter_organiser_submissions(city, api_key = Sys.getenv("AIRTABLE_API_KEY"))

Arguments

Value

A data frame with record_id, name, chapter, status, and type; empty when nothing matches.

Description

Finds open onboarding issues with no activity for days days and comments on each, naming the team that owns the first unchecked step rather than pinging everyone. Onboardings stall on the email and Meetup Pro teams more than anywhere else, so the ping is targeted.

Usage

chapter_remind_stale(
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding",
  days = 14,
  labels = c("new chapter", "chapter update")
)

Arguments

Value

Invisibly, a list of the issues nudged.

Description

Each chapter may carry a social_media$github value. Two shapes are legitimate: owner/repo for a chapter that keeps its files in a repository, and a bare organisation name for a chapter that runs its own GitHub organisation. Anything that resolves to neither is a dead link on the chapter's page.

Usage

chapter_repo_audit(chapters_dir, org = "rladies")

Arguments

Details

Where a reference is dead and the conventional ⁠<org>/meetup-presentations_<city>⁠ repository exists, that is reported as the likely correction.

Value

A data frame with columns file, github, state and suggestion. state is "repo", "org", "url", "prefixed" or "missing". A "url" row resolves but is written as a full URL; a "prefixed" row is an organisation the chapter owns, written as though it were a repository under org. Both carry the same reference in the expected shape as their suggestion. Only rows needing attention are returned.

Description

Render a chapter GitHub audit as markdown

Usage

chapter_repo_audit_report(audit)

Arguments

Value

A markdown string.

Description

Opt-in: not every chapter wants one, and only 111 of the 122 existing teams have a repository, so this is never called as part of onboarding without someone asking for it.

Usage

chapter_repo_create(team_slug, city, country, region = NULL, org = "rladies")

Arguments

Details

The repository is public and the chapter's team is given admin, matching every existing chapter repository - organisers administer their own.

Value

The repository name (invisibly).

Description

⁠meetup-presentations_<team slug>⁠, matching the 127 repositories the org already has. Hyphens are kept: san-diego, sao-paulo and kansas-city all have them, and sanfrancisco is the outlier rather than the rule.

Usage

chapter_repo_name(team_slug)

Arguments

Value

The repository name.

Description

Analyzes chapter activity data and returns a formatted markdown summary.

Usage

chapter_report_health(months = 6, org = "rladies")

Arguments

Value

Markdown report body (invisibly), or NULL if no data.

Description

Prompt the onboarding issue for the organiser Slack invite

Usage

chapter_slack_prompt(
  issue_number,
  city = NULL,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

The posted body (invisibly).

Description

Record that the organisers Slack invite was sent

Usage

chapter_slack_sent(
  issue_number,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

TRUE when the checklist step was ticked.

Description

The website's status field is set by hand, so it drifts from what the chapters are actually doing. This compares it against the event history in meetup_archive and reports the two mismatches worth acting on: a chapter running events that is not marked active, and a chapter marked active that has gone quiet.

Usage

chapter_status_reconcile(
  chapters_dir,
  health = NULL,
  months = 6,
  retire_months = 12,
  initiated = NULL
)

Arguments

Details

The categories follow the guide: a chapter is active when it has run an event in the last months, and inactive when it has not.

Chapters create their events on Meetup even when they promote them elsewhere, precisely so activity can be tracked, so an absence of Meetup events is a real signal rather than an artefact of where a chapter advertises. That holds for every chapter, so one with no Meetup group at all is included here rather than skipped: having no group means it never began, or stopped before it started.

Value

A data frame with one row per chapter that has a urlname,

A data frame with one row per chapter, with columns file, urlname, stored, last_event, months_inactive and action. action is "promote to active", "flag quiet", "mark inactive", or NA when nothing needs doing.

Description

Render a status reconciliation as markdown

Usage

chapter_status_reconcile_report(reconciled, months = 6)

Arguments

Value

A markdown string.

Description

Accepts either an issue number or a city name, and renders the checklist summary that ⁠/jinx chapter-status⁠ replies with.

Usage

chapter_status_report(
  ref,
  org = "rladies",
  onboarding_repo = "new-chapters-onboarding"
)

Arguments

Value

Markdown body as a single string.

Description

Reports rather than repairs. Renaming a team changes its URL and where its notifications go, so which of these to act on is a human decision.

Usage

chapter_team_audit(
  org = "rladies",
  website_repo = "rladies.github.io",
  parent_team = "chapters",
  statuses = "active"
)

Arguments

Details

Each row is one of:

• ok - the team matches a chapter's urlname

• slug_mismatch - a chapter matches by city, but its urlname would give a different slug

• no_chapter - no chapter in the website data explains this team

• missing_team - an active chapter with a Meetup group and no team

Value

A data frame with finding, team, display, chapter, and expected_slug.

Description

Render a team audit as markdown

Usage

chapter_team_audit_report(audit)

Arguments

Value

Markdown body as a single string.

Description

Creates the team nested under the chapters parent, named so that its slug matches the chapter's Meetup urlname, with the chapter spelled out in the description.

Usage

chapter_team_create(
  urlname,
  city,
  country,
  region = NULL,
  org = "rladies",
  parent_team = "chapters",
  overrides = character(0)
)

Arguments

Value

The created team's slug (invisibly).

Description

Free text, so unlike the name it can spell the chapter out in full without changing the slug.

Usage

chapter_team_description(city, country, region = NULL)

Arguments

Value

The description.

Description

GitHub derives a team's slug from its name and gives no way to set the slug directly, so the name has to be chosen such that it slugifies back to the chapter's urlname. That rules out "Portland, Oregon, USA" as a name - it would produce portland-oregon-usa rather than pdx.

Usage

chapter_team_name(slug, acronyms = chapter_team_acronyms())

Arguments

Details

The readable description of the chapter goes in the team's description instead, which is free text and does not affect the slug.

Value

The display name.

Description

The Meetup urlname, minus the ⁠rladies-⁠ prefix. This is the convention 108 of the 122 existing chapter teams already follow, and it is the only chapter field that is unique: seven city names are shared by more than one chapter, and it is the urlname that separates them - london from ldnont, pdx from portland-maine.

Usage

chapter_team_slug(urlname, overrides = character(0))

Arguments

Value

The team slug, or NA_character_ when there is no urlname.

Description

Runs the checks that keep ⁠data/chapters/*.json⁠ on the website consistent with the conventions jinx relies on elsewhere: the schema, the filename derivation, and the agreement between urlname and social_media$meetup.

Usage

chapter_validate_files(paths, added = character())

Arguments

Details

Designed to run over the files changed in a pull request, so that pre-existing drift in untouched files does not block unrelated work.

Value

A data frame with columns file, level ("error" or "warning"), check, and message. Zero rows when everything passes.

Description

Render chapter validation issues as markdown

Usage

chapter_validate_report(issues, n_files = NA_integer_)

Arguments

Value

A markdown string suitable for a pull request comment.

Description

Splits a markdown document by H1/H2 sections, then packs paragraphs into chunks of roughly target_chars characters. Returns a list of chunk records with heading, title, and lastmod metadata.

Usage

chunk_markdown(markdown, meta, target_chars = 1800, min_chars = 200)

Arguments

Value

List of chunk records (one list per chunk).

Description

Fails if the token has access to zero or more than one account.

Usage

cloudflare_account_id(api_token)

Arguments

Value

Account ID string.

Description

Calls the Cloudflare REST endpoint accounts/{id}/ai/run/{model} and returns the embedding vectors. Custom because cloudflarer does not wrap Workers AI inference; built on its cloudflarer::cf_request() and cloudflarer::cf_resp() for consistent auth and error handling.

Usage

cloudflare_embed(
  texts,
  account_id,
  api_token,
  model = "@cf/baai/bge-base-en-v1.5"
)

Arguments

Value

List of numeric vectors, one per input text.

Description

Calls the Cloudflare REST endpoint accounts/{id}/ai/run/{model} with a chat-style messages array and returns the generated text. Sibling of cloudflare_embed() - same auth/retry/error-handling idiom, built on cloudflarer::cf_request() and cloudflarer::cf_resp() because cloudflarer does not wrap Workers AI inference.

Usage

cloudflare_generate(
  messages,
  account_id,
  api_token,
  model = workers_ai_chat_model(),
  max_tokens = 256
)

Arguments

Value

Character scalar with the model's response text.

Description

Custom because cloudflarer does not wrap Vectorize v2; built on its cloudflarer::cf_request() and cloudflarer::cf_resp() for consistent auth and error handling.

Usage

cloudflare_vectorize_upsert(vectors, account_id, api_token, index_name)

Arguments

Value

The unwrapped result payload from the Cloudflare response.

Description

A handful of dispatched commands ("feedback", "setup-channel") act on the Slack workspace/channel the command was run from, rather than only on fields parsed from the command text itself. Owning that merge here - mirroring how event_parse() owns shaping the slack-event payload - keeps the calling workflow YAML a thin, untested plumbing layer instead of the place command semantics live.

Usage

cmd_attach_slack_context(command, team_id, channel_id, channel_name)

Arguments

Value

command with team_id/channel_id/channel_name added, or NULL unchanged if command was NULL.

Description

Privileged commands (those with a jinx_gated keyword in the command registry, i.e. anything not labelled jinx_safe) may only be run by members of the global team, identified via the Airtable member directory. Read-only commands are always allowed. The check fails closed: an unknown actor is denied, and a directory lookup error is also denied (with a distinct "try again" message).

Usage

cmd_authorize(
  command,
  actor = NULL,
  source = c("github", "slack"),
  workspace = NULL,
  api_key = Sys.getenv("AIRTABLE_API_KEY")
)

Arguments

Details

Slack identity is only trusted in the organisers workspace. The community workspace is openly joinable and its user ids come from a different Slack team, so an id from there must not authorize privileged actions; privileged Slack commands are only honored when workspace is "organiser".

Value

A list with ok (logical) and message (a refusal string when ok is FALSE, otherwise NULL).

Description

Returns the response message as a character string. The caller is responsible for routing the message to the right destination (GitHub issue comment, Slack, R console, etc.).

Usage

cmd_execute(command)

Arguments

Value

Character string with the response message.

Description

Parses commands like:

• ⁠/jinx invite @username to website⁠

• ⁠/jinx offboard @username from blog⁠

• ⁠/jinx report weekly⁠

• ⁠/jinx help⁠

Usage

cmd_parse(body)

Arguments

Value

A named list with action and action-specific fields, or NULL if the comment is not a jinx command.

Description

Summarizes active CFPs, speaker recommendations, and upcoming deadlines.

Usage

conference_generate_report(org = "rladies", repo = "global-team")

Arguments

Value

Formatted markdown string.

Description

List speaker recommendations for a conference

Usage

conference_list_speakers(conference, org = "rladies", repo = "global-team")

Arguments

Value

Data frame with columns: speaker, expertise, recommended_by.

Description

Creates a markdown table or image grid of contributors for inclusion in README or other documents.

Usage

contributor_format(contributors, format = c("table", "grid"), cols = 7)

Arguments

Value

Character string with markdown content.

Description

Collects all contributors from merged PRs and commits, returning a deduplicated list with contribution counts.

Usage

contributor_list(owner, repo, include_bots = FALSE)

Arguments

Value

Data frame with login, contributions, avatar_url, and profile_url columns, sorted by contribution count.

Description

Aggregates contributor data across all non-meetup repos in the org.

Usage

contributor_list_org(org = "rladies", exclude_pattern = "^meetup-")

Arguments

Value

Data frame with login, repos, contributions, avatar_url, and profile_url. contributions is the sum across all repos; repos is the count of distinct repos contributed to.

Description

Fetches contributors and commits the rendered list directly to the default branch. No PR is opened.

Usage

contributor_update(
  owner,
  repo,
  file_path = ".github/contributors.md",
  format = "grid",
  branch = "main"
)

Arguments

Value

Commit URL if changes were made, NULL otherwise (invisibly).

Description

Request GitHub Copilot as a reviewer on a pull request

Usage

copilot_request_review(
  owner,
  repo,
  pr_number,
  reviewer = copilot_reviewer_login()
)

Arguments

Value

TRUE if the request was accepted, FALSE otherwise.

Description

Requests Copilot as a reviewer and posts a short scoping comment naming the grimoire gate. Copilot then applies the instruction files synced by copilot_sync_repo(). Returns a status message suitable for replying to the requester.

Usage

copilot_review_pr(
  owner,
  repo,
  pr_number,
  gate = NULL,
  comment = TRUE,
  config = load_copilot_review_config()
)

Arguments

Value

A status string.

Description

Fetches the four grimoire review-gate skills (plus the foundation brand ruleset) and lands them in owner/repo as .github/copilot-instructions.md and path-scoped ⁠.github/instructions/*.instructions.md⁠ files, via a pull request. This is what makes GitHub Copilot's code review grimoire-aware. Re-run whenever grimoire changes to refresh the instructions.

Usage

copilot_sync_repo(
  owner = "rladies",
  repo,
  base = "main",
  branch = "jinx/copilot-review-instructions",
  config = load_copilot_review_config()
)

Arguments

Value

A list with status ("created" or "unchanged") and url (the PR URL when created).

Description

Uses magick to crop to a square and resize for web display.

Usage

directory_crop_image(
  path,
  width = 400,
  height = 400,
  gravity = "Center",
  output = path
)

Arguments

Value

Output path (invisibly).

Description

Reconciles Airtable against the directory and flips the directory_synced_field() checkbox on every submission whose work is done, so future syncs skip it:

Usage

directory_mark_synced(
  base_id = directory_base_id(),
  api_key = Sys.getenv("AIRTABLE_API_KEY"),
  org = "rladies",
  directory_repo = "directory"
)

Arguments

Details

• an update is done once applying it to main would change nothing — its entry is already fully incorporated (the same predicate the sync uses to decide whether to commit);

• a delete request is done once its entry file is absent from main — the reviewed purge workflow has removed it.

Stateless and idempotent — safe to run on every sync-PR merge and after a purge completes.

Value

Character vector of marked record ids (invisibly).

Description

Resizes to max width and compresses.

Usage

directory_optimize_image(path, max_width = 800, quality = 85, output = path)

Arguments

Value

Output path (invisibly).

Description

GDPR right-to-erasure: after the purge workflow removes a member's entry from the directory, every submissions row that resolves to one of slugs (by directory_id, falling back to identifier) is deleted from Airtable so no submitted PII remains — and so a later sync cannot re-create the entry from a leftover row. Unlike directory_mark_synced(), which only flags the synced checkbox, this destroys the rows.

Usage

directory_purge_submissions(
  slugs,
  base_id = directory_base_id(),
  api_key = Sys.getenv("AIRTABLE_API_KEY")
)

Arguments

Value

Character vector of deleted record ids (invisibly).

Description

Fetches directory submissions from Airtable, transforms each eligible record into the directory entry schema, and opens (or updates) a PR on the directory repo containing only the entries whose content changed.

Usage

directory_sync_airtable(
  base_id = directory_base_id(),
  api_key = Sys.getenv("AIRTABLE_API_KEY"),
  org = "rladies",
  directory_repo = "directory"
)

Arguments

Details

Only submissions marked minority_gender == "yes" and not flagged as a ⁠DEFAULT ROW⁠ are processed. Returning submitters are matched to their existing entry by slug (directory_id, falling back to identifier), and their submission is merged onto the existing file: fields listed in clear_fields are wiped first, then submitted fields overlay the rest, so a partial update never drops data the submitter left blank.

Delete requests are collected and reported in the PR body but are not executed here; destructive removal stays with the reviewed purge workflow.

Value

PR URL if changes found, NULL otherwise (invisibly).

Description

Checks that filenames follow conventions: lowercase, no hashes, ASCII-only, .json extension.

Usage

directory_validate_filename(filename)

Arguments

Value

A named list with valid (logical) and issues (character vector of problems found).

Description

Performs HTTP HEAD requests to check if profiles resolve.

Usage

directory_verify_handles(entry)

Arguments

Value

A data frame with columns platform, handle, status, and valid.

Description

Events are gated on the originating team being an allowed workspace, not on actor privilege (there is no "actor" for a passive event like a reaction or a team join - anyone triggers these). airtable_webhook events have no team_id at all (Airtable webhooks aren't Slack-team scoped) and are trusted by construction: the Worker already verified the webhook's shared secret before dispatching. The other three kinds are checked against the organiser/community team allowlist, mirroring the Worker's own slack_team_is_allowed().

Usage

event_authorize(
  event,
  organiser_id = Sys.getenv("SLACK_ORGANIZER_TEAM_ID"),
  community_id = Sys.getenv("SLACK_COMMUNITY_TEAM_ID")
)

Arguments

Value

A list with ok (logical) and message (a refusal string when ok is FALSE, otherwise NULL).

Description

Create a formatted event summary

Usage

event_create_summary(events, period = c("weekly", "monthly"))

Arguments

Value

Formatted markdown string.

Description

Unlike cmd_execute(), handlers perform their own Slack/Airtable API calls directly rather than returning a reply string for a workflow step to relay - there is no single "response destination" for a passive event the way there is for a command. The return value is a short status string for the GitHub Actions run log only.

Usage

event_execute(event)

Arguments

Value

Character string describing the outcome, for logging.

Description

Queries Meetup Pro for recent events.

Usage

event_list_chapter(chapter, months = 3)

Arguments

Value

Data frame with columns: title, date, url, rsvp_count, source, chapter.

Description

The Worker's repository_dispatch payload is already structured JSON (unlike slash commands, which arrive as free text for cmd_parse() to split), so this is a thin validator rather than a parser: it confirms kind is one of the registered jinx_events() entries and normalizes the shape callers can rely on.

Usage

event_parse(payload)

Arguments

Value

A list with kind, team_id, response_url, event; kind is "unknown" when the payload's kind isn't registered.

Description

Fetches events from all configured sources and writes a summary.

Usage

event_sync_chapters(
  org = "rladies",
  target_repo = "event-archive",
  months = 3,
  dry_run = TRUE,
  chapters = NULL
)

Arguments

Details

The chapter list comes from the website's data/chapters entries via chapter_meetup_groups(), not from configuration: that data is already the source of truth for which chapters exist and which Meetup group each uses, and a hand-maintained list would drift immediately.

Value

Data frame of all events (invisibly).

Description

Looks up ⁠gather_<type>()⁠ where ⁠<type>⁠ has dashes replaced by underscores. Sources extend the indexer by defining a new ⁠gather_<type>(src)⁠ function.

Usage

gather_rag_source(src)

Arguments

Value

List of chunk records (with chunk_idx set per document).

Description

Reads the base branch tip and either creates branch there or force-updates it. Used by helpers that need to (re-)open a clean working branch before pushing files.

Usage

gh_branch_upsert(org, repo, branch, base = "main", force = TRUE)

Arguments

Value

The SHA the branch now points to (invisibly).

Description

Thin wrapper around gh_welcome_contributor() that fixes is_pr = TRUE for use in PR-only workflows. Reusable across any repo.

Usage

gh_greet_contributor(
  owner,
  repo,
  number,
  author,
  org = "rladies",
  extra_message = NULL
)

Arguments

Value

Comment URL or NULL if author is a team member (invisibly).

Description

If a PR is already open from branch into base, returns its URL; otherwise opens a new PR with the given title and body.

Usage

gh_open_or_update_pr(
  org,
  repo,
  branch,
  base = "main",
  title,
  body,
  team_reviewers = NULL
)

Arguments

Value

PR URL.

Description

Posts a review checklist comment tailored for content PRs (blog posts, news posts, or any markdown-driven content). Generic across content kinds — the only difference between blog and news today is the path filter the caller workflow applies.

Usage

gh_post_checklist(owner, repo, pr_number)

Arguments

Value

Comment URL (invisibly).

Description

Posts a thank-you message tailored to whether this is the author's first merged PR, and credits everyone else who helped: co-authors (commit authors and ⁠Co-authored-by:⁠ trailers), reviewers, and people who commented on the PR.

Usage

gh_thank_contributor(owner, repo, pr_number, author, org = "rladies")

Arguments

Value

Comment URL (invisibly).

Description

Posts a thank-you when an issue is closed, crediting the person who raised it (with a warmer message if it was their first here) plus everyone who was assigned to it or commented on it.

Usage

gh_thank_issue_contributors(owner, repo, number, author, completed = TRUE)

Arguments

Value

Comment URL (invisibly), or NULL for bot authors.

Description

Checks if the PR/issue author has previous contributions to the repo. If this is their first, posts a warm welcome message. If not, posts a shorter thank-you. No-op for bots and global team members.

Usage

gh_welcome_contributor(
  owner,
  repo,
  number,
  author,
  is_pr = TRUE,
  org = "rladies",
  extra_message = NULL
)

Arguments

Value

Comment URL or NULL if author is a team member (invisibly).

Description

Scans all non-meetup repositories in the organization, collects workflow information, and generates a JSON report.

Usage

gha_generate_dashboard(
  org = "rladies",
  exclude_pattern = "^meetup-",
  output_path = NULL
)

Arguments

Value

List of workflow data (invisibly).

Description

Lists pending invitations and triggers finalization for accepted ones.

Usage

gt_check_invitations(org = "rladies")

Arguments

Value

Character vector of usernames that have accepted their invitation (invisibly).

Description

Create a global team offboarding issue

Usage

gt_create_offboarding(username, team, name = username, org = "rladies")

Arguments

Value

The created issue URL (invisibly).

Description

Finalize global team offboarding by removing user from teams

Usage

gt_finalize_offboarding(username, team, org = "rladies")

Arguments

Value

Invisibly returns NULL. Called for its side effect of removing the user from the specified teams.

Description

Adds the user to their specific team and creates a tracking issue with the onboarding checklist.

Usage

gt_finalize_onboarding(username, team, name = username, org = "rladies")

Arguments

Value

The created issue URL (invisibly).

Description

Sends an organization membership invitation and adds the user to the global team.

Usage

gt_invite(username, team, name = username, org = "rladies")

Arguments

Value

Invisibly returns NULL. Called for its side effect of sending the invitation and adding the user to the specified team.

Description

Finds open issues with onboarding or offboarding labels that haven't been updated in days days, and posts a reminder comment.

Usage

gt_remind_stale(org = "rladies", days = 30, repo = "global-team")

Arguments

Value

Invisibly returns NULL. Called for its side effect of posting reminder comments on stale issues.

Description

Part of global team onboarding: opens a "find a time" poll on samkoma so the new member and their team can settle on an onboarding meeting slot. The poll offers a two-week window of candidate dates starting one week after it is created, and the poll link is posted as a comment on the onboarding issue. The poll is created (and the comment posted) under Jinx's identity.

Usage

gt_schedule_onboarding_meeting(
  issue_number,
  name,
  team_name,
  org = "rladies",
  repo = "global-team",
  start = Sys.Date() + 7,
  from = "08:00",
  to = "20:00",
  slot = 30,
  tz = "UTC"
)

Arguments

Value

The poll URL (invisibly).

Description

The guide writes the templates that are sent as email with their subject on a leading ⁠SUBJECT:⁠ line, followed by ⁠BODY OF THE MESSAGE:⁠. This splits the two apart so that callers send the guide's subject rather than inventing one, and errors if the template carries no subject at all.

Usage

guide_email_template(name, variables = list(), base_url = guide_base_url())

Arguments

Value

A list with name, subject and body.

See Also

guide_template() for a template with no subject line.

Examples

## Not run: 
notice <- guide_email_template("chapter-inactive-first-notice")
notice$subject

## End(Not run)

Description

Organisational communications - what RLadies+ says to organisers and chapters - are written and edited in the guide, under ⁠static/templates/⁠, and served as markdown from ⁠https://guide.rladies.org/templates/<name>.md⁠. jinx fetches them instead of shipping copies, so a volunteer can fix the wording without waiting for a package release, and the text an organiser reads in the guide is the text jinx sends.

Usage

guide_template(name, variables = list(), base_url = guide_base_url())

Arguments

Details

A failed fetch is an error. There is deliberately no vendored fallback, because a second copy is the drift this replaces.

Use guide_email_template() for a template that is sent as email and so carries its own subject line.

Value

The template text, as a single string.

Examples

## Not run: 
guide_template("meetup-group-description")

guide_template(
  "chapter-onboarding-welcome",
  list(FIRST_NAME = "Ada", CITY = "Oslo")
)

## End(Not run)

Description

Check translation coverage across languages

Usage

i18n_check_coverage()

Value

Data frame with columns: language, total_templates, translated, coverage_pct.

Description

Looks up the chapter's language preference from its metadata. Falls back to English if not set.

Usage

i18n_get_chapter_language(chapter, org = "rladies")

Arguments

Value

Language code string.

Description

List supported languages

Usage

i18n_list_languages()

Value

Data frame with columns: code, name, native_name, direction.

Description

Looks up ⁠inst/translations/{language}/{template_name}.md⁠ first, falls back to ⁠inst/templates/{template_name}.md⁠ if the translation is missing.

Usage

i18n_translate_template(template_name, language = "en", variables = list())

Arguments

Value

Rendered template as a single character string.

Description

Checks that all translated templates have the same ⁠<KEY>⁠ placeholders as the English baseline.

Usage

i18n_validate_translations(language = NULL)

Arguments

Value

Data frame with columns: template, language, status, missing_keys, extra_keys.

Description

Get LinkedIn API version string

Usage

li_get_version()

Value

Version string in YYYYMM format.

Description

Upload media to LinkedIn

Usage

li_media_upload(author, media)

Arguments