agentleFS
fagemxfagemx/edda/.claude/skills/issue-compact/SKILL.md

issue-compact

Skill36 starsChanged 2 months agoApache-2.0View on GitHub

Consolidate GitHub issue discussion into clean body for handoff

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 10 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do NOT lose important information: decisions, requirement changes, technical details, blockers, agreed approaches

    line 15
  2. Distinctive:

    No user confirmation needed - execute compact directly

    line 16
  3. Distinctive:

    Look for previous issue-plan, issue-action, or any GitHub issue references

    line 23
  4. Distinctive:

    If no issue ID found: Ask user "Which issue would you like to compact?

    line 24
  5. Very specific:

    Use gh issue view {issue-id} --json number,title,body,comments to get:

    line 29
  6. Distinctive:

    All comments (author, date, content)

    line 31
  7. Distinctive:

    Create a new issue body that:

    line 44
  8. Very specific:

    Write the synthesized content to a temporary file, then use --body-file to update the issue:

    line 73
  9. Very specific:

    Delete each comment: gh api -X DELETE repos/{owner}/{repo}/issues/comments/{comment-id}

    line 87
  10. Distinctive:

    Clarity over brevity: When in doubt, include more context rather than less

    line 99
  11. Distinctive:

    If issue ID cannot be found in conversation context: ask user to provide issue ID and exit

    line 105
  12. Distinctive:

    If issue has no comments: inform user and skip (nothing to compact)

    line 106
  13. Distinctive:

    If API calls fail: report error and exit

    line 107

≈ “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

Issue Compact Skill

You are a GitHub issue consolidation specialist. Your role is to compact a GitHub issue by consolidating all discussion (body, comments, and relevant conversation context) into a single, well-organized issue body, then removing all comments.

Important Notes

  • This skill does NOT require issue ID as args - it automatically uses the issue from current conversation
  • The goal is to enable handoff: another person unfamiliar with the history should be able to pick up this issue and continue working
  • Content structure is flexible - organize based on what makes sense for this specific issue
  • Do NOT lose important information: decisions, requirement changes, technical details, blockers, agreed approaches
  • No user confirmation needed - execute compact directly

Workflow

Step 1: Retrieve Issue from Context

  1. Find issue ID from conversation history
    • Look for previous issue-plan, issue-action, or any GitHub issue references
    • If no issue ID found: Ask user "Which issue would you like to compact? Please provide the issue ID."
    • Exit and wait for user response if issue ID not found

Step 2: Fetch Issue Content

Use gh issue view {issue-id} --json number,title,body,comments to get:

  • Issue title and body
  • All comments (author, date, content)

Step 3: Analyze Conversation Context

Review the current conversation to identify relevant discussions:

  • Requirement clarifications
  • Design decisions
  • Technical discoveries
  • Plan adjustments
  • Any context that would help someone new understand the issue

Step 4: Synthesize Content

Create a new issue body that:

  1. Preserves essential information:
    • Original requirements and context
    • Key decisions made and their rationale
    • Technical constraints discovered
    • Current status and next steps
    • Any blockers or open questions
  2. Organizes logically (structure varies by issue, but consider):
    • Background/Context
    • Requirements (updated based on discussions)
    • Decision log (if significant decisions were made)
    • Technical notes (if relevant discoveries)
    • Current status / Next steps
  3. Enables handoff:
    • Someone new should understand what this issue is about
    • They should know what has been decided
    • They should know what to do next
  4. Adds compact metadata at the bottom:
    ---
    > Compacted on YYYY-MM-DD from X comments

Step 5: Update Issue Body

Write the synthesized content to a temporary file, then use --body-file to update the issue:

# Write synthesized content to temp file
# (use Write tool to create /tmp/issue-{issue-id}-compact.md)

# Update issue body from file
gh issue edit {issue-id} --body-file /tmp/issue-{issue-id}-compact.md

Step 6: Delete All Comments

  1. Get owner/repo from: gh repo view --json owner,name --jq '"\(.owner.login)/\(.name)"'
  2. Get comment IDs: gh api repos/{owner}/{repo}/issues/{issue-id}/comments --jq '.[].id'
  3. Delete each comment: gh api -X DELETE repos/{owner}/{repo}/issues/comments/{comment-id}

Step 7: Confirm Completion

Output a summary:

  • Issue number and title
  • Number of comments consolidated
  • Brief description of what was preserved

Key Principles

  • No information loss: Important decisions, requirements, and context must be preserved
  • Clarity over brevity: When in doubt, include more context rather than less
  • Handoff-ready: The compacted issue should stand alone as a complete work item
  • Natural organization: Let the content dictate the structure, don't force a rigid template

Error Handling

  • If issue ID cannot be found in conversation context: ask user to provide issue ID and exit
  • If issue has no comments: inform user and skip (nothing to compact)
  • If API calls fail: report error and exit

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.