agentleFS
MarcusJellinghausMarcusJellinghaus/mcp-workspace/.claude/skills/issue_approve/SKILL.md

issue_approve

Skill49 starsChanged 2 days agoMITView on GitHub

Approve issue to transition to next workflow status

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 5 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If no issue number is provided:

    line 21
  2. Very specific:

    Check if the issue number is known from prior /issue_analyse or /issue_create in this conversation

    line 22
  3. Very specific:

    Read .vscodeclaude_status.txt and extract the issue number from the Issue #NNN line

    line 23
  4. Distinctive:

    If still unknown, ask the user

    line 24
  5. Very specific:

    If --repo owner/repo was given, compare it with origin (mcp__mcp-workspace__git, command ls_remote, args ["--get-url", "origin"]).

    line 28
  6. Very specific:

    No match (or no origin) → cross-repo.

    line 33
  7. Distinctive:

    No matching entry → stop and report "not a reference project".

    line 36
  8. Very specific:

    Call mcp__mcp-workspace__github_issue_view with the issue number (plus reference_name for cross-repo).

    line 42
  9. Distinctive:

    No blocking questions remain

    line 53
  10. Distinctive:

    If any check fails, stop and report — do not approve.

    line 55
  11. Very specific:

    Do not also comment /approve — the Action would race this call.

    line 78

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Approve Issue

Approve the current issue to transition it to the next status in the workflow.

Resolve Issue Number

The user may provide an issue number as the argument (available as $ARGUMENTS). If no issue number is provided:

  1. Check if the issue number is known from prior /issue_analyse or /issue_create in this conversation
  2. Read .vscodeclaude_status.txt and extract the issue number from the Issue #NNN line
  3. If still unknown, ask the user

Cross-Repo Issues

  • If --repo owner/repo was given, compare it with origin (mcp__mcp-workspace__git, command ls_remote, args ["--get-url", "origin"]). Normalise both: accept https://github.com/owner/repo and git@github.com:owner/repo forms, drop a trailing .git, compare case-insensitively.
  • Match → ignore --repo entirely; follow the normal (workspace) path.
  • No match (or no origin) → cross-repo. Call mcp__mcp-workspace__get_reference_projects() and pick the entry whose url normalises (as above) to owner/repo. Use its name as reference_name on every GitHub tool call. Entries with a null url never match.
  • No matching entry → stop and report "not a reference project". There is no other route.
  • Never pass a repo to set-status.

Instructions

  1. Fetch the issue to confirm it exists: Call mcp__mcp-workspace__github_issue_view with the issue number (plus reference_name for cross-repo). Cross-repo only: success means the output starts with # #<number>: , or is # #<number> followed by a newline (untrusted author; proceed normally). Anything else → "approval failed: <result>".

    Note its status labels (labels prefixed status-) — step 3 uses them for the promotable check and the --from-status value.

  2. Validate that the issue is ready for approval:

    • Issue has been analyzed/discussed
    • Requirements are clear
    • No blocking questions remain

    If any check fails, stop and report — do not approve.

  3. Promote.

    currentnext
    status-01:createdstatus-02:awaiting-planning
    status-04:plan-reviewstatus-05:plan-ready
    status-07:code-reviewstatus-08:ready-pr
    • More than one status label → don't promote; report "nothing to approve (multiple status labels: X, Y)".
    • Current status not in the table (including no status label) → don't promote; report "nothing to approve".
    • Cross-repo → call mcp__mcp-workspace__github_issue_comment(number=…, reference_name=…, body="/approve") and report "approval requested, not confirmed". No polling.
    • Otherwise run from the project directory (no --project-dir, no output redirection — the result is printed on stderr):
      mcp-coder gh-tool set-status <next> --from-status <current> --issue <issue_number> --force
      --force only skips the clean-tree check, irrelevant for a label change. Do not also comment /approve — the Action would race this call.
  4. Report the issue number and the outcome:

    set-status resultReport
    exit 0, Updated issue #N to <next><current> → <next>
    exit 0, Skipped set-status to '…': current label is 'X', expected 'Y'"not approved (current status X)"
    non-zero exit"approval failed: <error output>" — stop; no retry, no /approve fallback
    Cross-repo comment resultReport
    starts with Added comment to issue #"approval requested, not confirmed"
    anything else"approval failed: <result>" — no retry

    Not-called outcomes are reported as stated in step 3.

Note: This skill has disable-model-invocation — it can only be run by the user typing /issue_approve. If you need this skill as a follow-up, tell the user: "Please run /issue_approve to proceed."

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.