agentleFS
ByronScottJonesByronScottJones/Agents/skills/twg-jira-resolve-merged-work/SKILL.md

twg-jira-resolve-merged-work

Skill7 starsChanged 2 months agoMITView on GitHub

Clean up stale or unresolved Jira workitems. Dry-run single items, lists, boards, sprints, epics, or projects by matching Jira keys/titles to merged PRs, repos, Rovo/search-code hits, and assignee activity.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 8 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    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
  1. Very specific:

    Use exact command grammar from live twg help / twg help describe <path>.

    line 14
  2. Very specific:

    Run twg <command>. On shell command not found, use $HOME/.local/bin/twg (macOS/Linux) / $env:LOCALAPPDATA\Programs\twg\bin\twg.exe …

    line 19
  3. Distinctive:

    Default to dry run. Ask before mutating unless the user requests reviewed execution.

    line 34
  4. Distinctive:

    Use completed peers to infer common repos, assignees, Jira-key title/branch conventions, PR links, and merge-to-transition lag.

    line 58
  5. Distinctive:

    Check optional discovery signals.

    line 61
  6. Very specific:

    If rovo list-apps shows Bitbucket or GitHub connected, use Rovo search for keys, titles, PR URLs, branches, and repos.

    line 62
  7. Very specific:

    If search-code is available, search exact keys first, then distinctive title terms in discovered repos.

    line 65
  8. Distinctive:

    Prefer assignees from completed peer items and candidates.

    line 70
  9. Distinctive:

    Search merged PRs in batches.

    line 75
  10. Distinctive:

    Prefer one merged-PR query per repo/window, then match locally across all candidate keys.

    line 76
  11. Distinctive:

    Only mutate after explicit approval of the exact rows or plan.

    line 99
  12. Distinctive:

    Do not transition medium-confidence items automatically.

    line 113

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

twg-jira-resolve-merged-work

Resolve stale Jira workitems only when merged PR evidence is strong. Plan first; mutate only after approval.

Use exact command grammar from live twg help / twg help describe <path>; do not guess board, sprint, transition, workspace, repo, PR, or field syntax.

CLI launcher fallback

Run twg <command>. On shell command not found, use $HOME/.local/bin/twg (macOS/Linux) / $env:LOCALAPPDATA\Programs\twg\bin\twg.exe (PowerShell), then tell user to add that directory to PATH. Do not treat auth or command errors as PATH failures.

Scope

Accept one Jira scope:

  • One Jira workitem key/URL.
  • A small explicit list of Jira workitem keys/URLs.
  • Board or sprint ID/URL.
  • Epic key.
  • Jira space/project key.

Default to dry run. Ask before mutating unless the user requests reviewed execution. Use a bounded window; otherwise inspect recent active/completed sprint work and nearby PR merge dates.

Workflow

  1. Load Jira candidates.
    • Single/list: hydrate each provided workitem directly.
    • Sprint: jira sprint workitems query.
    • Board: board/backlog/sprint commands advertised by live help.
    • Epic or space: JQL-backed jira workitem query.
    • Hydrate key, title, status/category, assignee, updated time, parent, subtasks/blockers, and URL.
  2. Split the set.
    • Candidates are not done/resolved. Peers are done in the same board, sprint, epic, or space.
    • Exclude epics/parents. Skip unresolved blockers or incomplete subtasks unless the user explicitly allows them.
  3. Discover shared repo context set-wise.
    • Even for one item or a short list, infer context for the set together.
    • Extract repos/workspaces from linked PRs, commits, branches, project links, Compass/components, context queries, and resolved peers.
    • Use completed peers to infer common repos, assignees, Jira-key title/branch conventions, PR links, and merge-to-transition lag.
  4. Check optional discovery signals.
    • If rovo list-apps shows Bitbucket or GitHub connected, use Rovo search for keys, titles, PR URLs, branches, and repos. Treat it as discovery until hydrated through TWG/provider/Jira evidence.
    • If search-code is available, search exact keys first, then distinctive title terms in discovered repos. Strong hits are branch names, commit text, changed files, or symbols; broad fuzzy hits stay weak.
  5. Expand likely implementers.
    • Prefer assignees from completed peer items and candidates. Treat reporters and commenters as weak hints.
    • Query merged PRs authored by these assignees in the window, scoped by discovered repos first; if unknown, infer repos from author/date results.
  6. Search merged PRs in batches.
    • Prefer one merged-PR query per repo/window, then match locally across all candidate keys. One PR may satisfy multiple exact keys.
    • Match exact keys in PR title, description, branch, commits, and linked issue metadata.
    • Fall back to per-workitem lookup only for no evidence, priority items, or ambiguity.
  7. Score confidence.
    • High: exact key in merged PR or linked PR metadata, plausible timing, no open same-key PR, no blockers/subtasks, and target transition exists.
    • Medium: same assignee/repo with strong title similarity, or exact key only in commit/search-code evidence.
    • Low: fuzzy similarity, same author only, unknown repo, or ambiguity.
  8. Produce the dry-run plan.
    • Table columns: workitem key/title, current status, proposed transition, confidence, brief rationale, PR title/URL, merge date, last Jira update, link/repo proposal, PR title-fix proposal, skipped reason, and verification.
    • Include exact commands only after live-help verification. Never present low-confidence rows as executable.

Mutations

Only mutate after explicit approval of the exact rows or plan.

  • Transition: list transitions first, then use only the approved done/resolved target.
  • Link enrichment: if links are missing, offer jira workitem link weblink for PR/repo URLs after querying links and verifying help. Do not claim this creates native development-panel links.
  • PR title fix: for Bitbucket, update title only when a verified bitbucket pull-requests get leaf can read the PR, write access is verified, the PR is confidently tied to the item, and the user approved it. For GitHub, update only when an authenticated write tool is available and access is verified; otherwise skip with the reason.
  • Verification: re-read each changed workitem and PR. Report changed keys, statuses, URLs, and failed rows.

Do not transition medium-confidence items automatically. Do not transition declined/superseded PRs or rows supported only by same author, same repo, or fuzzy similarity.

Output

Lead with the decision summary:

  • executable high-confidence rows,
  • review-needed medium-confidence rows,
  • skipped unsafe rows,
  • missing repo/auth/permission coverage.

Then provide the dry-run table and exact approval question. Include URLs or stable IDs for every workitem and PR used as evidence.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.