briefing
Skill5 starsChanged 44 days agoMITView on GitHub
User-invoked only. Read-only context build: the ticket, its comment trail, epic chain, and every PR raised against it
- Reads credentials
20 other projects use a version of this.
Adopted by
Most recently taken up by
- PatrickSUDO/fadacai-portfoliotheir own version, first seen Nov 13, 2026
- adeonir/agent-skillstheir own version, first seen Nov 10, 2026
- gustaferiksson/skillstheir own version, first seen Nov 8, 2026
- t128n/skillstheir own version, first seen Nov 8, 2026
- parker-brown-family/agent-skillstheir own version, first seen Nov 8, 2026
- Getty/briefingtheir own version, first seen Nov 8, 2026
Used with
Projects with it also use handoff and grill-me.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.Other versions20 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 20
- garrytan/gbrain★ 33ksame name
- PatrickSUDO/fadacai-portfolio★ 145same name
- google-gemini/gemini-managed-agents-templates★ 99same name
- norrietaylor/distillery★ 25same name
- bks-lab/open-bridge★ 13same name
- adeonir/agent-skills★ 8same name
- Getty/briefing★ 5same name
- Erionis/Jarvis-Lite★ 3same name
- flopperj/dotfiles-core★ 3same name
- alyan06/claude-skills★ 1same name
10 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository24 other files it gives its agents.
- AGENTS.mdAGENTS.md
- assign-epicSkill
- bump-versionSkill
- captureSkill
- consultantSkill
- dependabot-reviewSkill
- fix-cicdSkill
- log-timeSkill
- morlockSkill
- morningSkill
- multi-repoSkill
- nightly-checkSkill
- pickupSkill
- pr-action-review-mine-loopSkill
- pr-action-reviewSkill
- pr-choreSkill
- pr-review-loopSkill
- prSkill
- pushSkill
- qa-review-actionSkill
- refineSkill
- runSkill
- syncSkill
- wrap-upSkill
The file
We split this skill into the rules it gives an agent
10 rules. 0 are common, 6 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 17 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
Ticket input: use the value supplied with the skill invocation.
- Distinctive: line 13
Read-only. Does not assign, transition, comment on, or otherwise modify the ticket or any PR — this is purely for getting the human up to …
- Very specific: line 19
If the ticket input is a bare number (e.g.
- Very specific: line 23
Check that JIRA_BASE_URL, JIRA_API_TOKEN, JIRA_EMAIL are set in .env in the current repo.
- Very specific: line 50
Read every comment oldest-first — author.displayName, created, rendered body.
- Very specific: line 56
If fields.customfield_10014 is set and differs from the parent chain, fetch that too.
- Distinctive: line 64
Search across all PR states — open, merged, and closed-without-merge all matter for context:
- Distinctive: line 105
If a PR is large, read the diff file-by-file rather than skimming — the goal is to actually understand the shipped code, not just know that …
- Distinctive: line 130
Keep this tight — it exists so the user can start follow-on work without re-reading everything themselves, not to reproduce everything you …
- Distinctive: line 134
Do not take any action on the ticket or any PR.
“Common” is measured: kept by at least 6 in 10 of the 17 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Build full context on a Jira ticket before follow-on work: the ticket, its comments, its epic/parent chain, every PR raised against it, and the actual code those PRs shipped.
Ticket input: use the value supplied with the skill invocation.
Usage: briefing <ticket-id>
Read-only. Does not assign, transition, comment on, or otherwise modify the ticket or any PR — this is purely for getting the human up to speed.
Step 0 — Normalise the ticket ID
If the ticket input is a bare number (e.g. 25), prepend $JIRA_PROJECT_KEY- to get e.g. PROJ-25. Store the result as TICKET_KEY.
Step 1 — Verify credentials
Check that JIRA_BASE_URL, JIRA_API_TOKEN, JIRA_EMAIL are set in .env in the current repo. If missing, stop and ask the user to add them. Source .env before all Jira calls. Confirm gh auth status succeeds before any GitHub calls.
Step 2 — Fetch the ticket and its full comment thread
source .env && curl -s -u "$JIRA_EMAIL:$JIRA_API_TOKEN" \
"$JIRA_BASE_URL/rest/api/3/issue/$TICKET_KEY?expand=renderedFields"Extract fields.summary, fields.status.name, fields.issuetype.name, fields.parent,
fields.customfield_10014 (epic link), fields.subtasks, fields.issuelinks, fields.labels.
For the description, read renderedFields.description — that's the HTML the expand=renderedFields
above returns. On API v3 fields.description is Atlassian Document Format (a nested JSON node tree),
which is why the expand is there: it saves you walking ADF to recover the prose.
source .env && curl -s -u "$JIRA_EMAIL:$JIRA_API_TOKEN" \
"$JIRA_BASE_URL/rest/api/3/issue/$TICKET_KEY/comment?orderBy=created&expand=renderedBody"Comment bodies are ADF too, so expand=renderedBody is what makes renderedBody (HTML) available
alongside the raw body — read the rendered form. The endpoint is paginated: compare total against
startAt + maxResults and re-request with a higher startAt until you have them all. A long-running
ticket is exactly the case where the early comments matter most, and they're the ones a single
unpaginated call drops.
Read every comment oldest-first — author.displayName, created, rendered body. This is the decision trail: who asked what, who answered, what changed direction mid-ticket. Distinguish an agent's own progress summary from genuine human input — the former restates what was done, the latter is a decision.
Step 3 — Walk the epic/parent chain
If fields.parent is set, fetch it. If that parent also has a parent, keep walking up until you reach the epic (issuetype.name == "Epic") or run out of ancestors. Fetch the epic itself and read its description and comments the same way as Step 2 — the epic is where the ticket's why lives, not just its what.
If fields.customfield_10014 is set and differs from the parent chain, fetch that too.
Step 4 — Fetch subtasks and linked issues
For each key in fields.subtasks, fetch the full issue and its comments. For each entry in fields.issuelinks, note the link type (blocks/relates to/etc.) and fetch a summary of the linked issue — enough to know if it changes how the ticket should be read, not a full deep-dive unless it looks load-bearing.
Step 5 — Find every PR raised against the ticket
Search across all PR states — open, merged, and closed-without-merge all matter for context:
gh pr list --state all --search "$TICKET_KEY" \
--json number,title,state,url,body,mergedAt,author,createdAt--search only covers what GitHub indexes for the PR itself, so sweep the full PR list directly as a
second pass — matching the key against the title, the head branch name, and the body:
gh api "repos/{owner}/{repo}/pulls?state=all" --paginate \
--jq ".[] | select((.title, .head.ref, (.body // \"\")) | test(\"$TICKET_KEY\"; \"i\")) | {number, title, state}"Pass state in the query string, not as -f state=all — gh api switches to POST as soon as a field
is supplied, which would attempt to open a pull request instead of listing them.
Neither pass sees inside commit messages. If you suspect a PR carries the ticket only in its commits — a renamed or stacked branch is the usual cause — check the candidates explicitly:
gh pr view <number> --json commits --jq '.commits[].messageHeadline'Build the full list of PR numbers touching this ticket before moving on — don't process them one at a time as you find them, since a later search may surface one you'd already started summarising without it.
Step 6 — Read each PR: description, discussion, and diff
For every PR number found:
gh pr view <number> --json title,body,state,mergedAt,files,additions,deletions
gh pr diff <number>
gh api repos/{owner}/{repo}/pulls/<number>/comments --paginate
gh api repos/{owner}/{repo}/issues/<number>/comments --paginate
gh api repos/{owner}/{repo}/pulls/<number>/reviews --paginateFor each PR, note: what it actually changed (from the diff, not just the title), why (from the description), what reviewers pushed back on or asked to change, and whether feedback was accepted or argued down. A PR that was closed without merging is still context — it tells you an approach was tried and abandoned, and usually why.
If a PR is large, read the diff file-by-file rather than skimming — the goal is to actually understand the shipped code, not just know that code exists.
Step 7 — Synthesise
Produce a single narrative brief, in this order:
## Briefing: <ticket-id> — <summary>
### The ask
What the ticket (and epic, if it adds meaning) actually asks for, in your own words — not a copy of the description.
### Timeline
Chronological walk through: ticket created → key comments/decisions → PR(s) raised → review discussion → merge (or abandonment). Call out any direction changes and why they happened.
### What shipped
Per merged PR: a short summary of the actual code change (files/areas touched, the approach taken), not just "implemented the feature."
### Decisions and constraints established
Anything settled during discussion that isn't obvious from the ticket text alone — naming choices, scope cuts, deliberate deferrals, "we decided not to do X because Y."
### Loose ends
Anything left open: follow-up tickets mentioned, known limitations acknowledged in review, TODOs left in the shipped code, unresolved PR comments.Keep this tight — it exists so the user can start follow-on work without re-reading everything themselves, not to reproduce everything you read.
Step 8 — Hand back
Do not take any action on the ticket or any PR. End by asking the user what they want to do with this context — the brief is preparation, not a deliverable in itself.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

