agentleFS
MicrosoftKnown publishermicrosoft/IssueLens/skills/assign-issue/SKILL.md

assign-issue

Skill0 starsChanged 24 days agoMITView on GitHub

Assign GitHub issues to the right person based on technical area ownership and historical patterns. Use when auto-assigning new issues, finding the right owner, or routing issues to area experts. Triggers on "assign issue", "find owner for issue", "who should handle this issue", and "route this issue".

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. About 0 are common, 12 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    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. Very specific:

    Follow the issuelens-config skill and call the issuelens-config tool for domain assignment.

    line 29
  2. Very specific:

    If the source is built-in, no static mapping exists and historical patterns are the primary strategy.

    line 42
  3. Distinctive:

    Use the available GitHub issue-search operation to find similar closed (and recently resolved) issues in the same repository:

    line 48
  4. Distinctive:

    Search using the most specific terms from the issue title, body, and labels.

    line 51
  5. Distinctive:

    Prefer issues updated in the last 6 to 12 months.

    line 52
  6. Distinctive:

    Do not infer ownership from issue authorship alone.

    line 57
  7. Distinctive:

    Read the target issue and preserve its existing assignees.

    line 67
  8. Distinctive:

    Read the issue again and confirm that the selected assignee is present.

    line 81
  9. Very specific:

    Use only the bundled IssueLens GitHub MCP tools for every GitHub read or write.

    line 86
  10. Distinctive:

    Never use a shell command, bundled script, direct HTTP request, or GitHub CLI command.

    line 88
  11. Distinctive:

    Never remove or replace existing assignees.

    line 90
  12. Distinctive:

    Do not assign a team handle directly.

    line 91
  13. Distinctive:

    Treat a failed or unconfirmed write as a suggestion, not a completed action.

    line 95
  14. Distinctive:

    If no area-owner file exists, use historical patterns as the primary strategy: search similar past issues and assign the most frequent / …

    line 96
  15. Distinctive:

    If no confident match exists, do not guess and do not modify the issue.

    line 100

≈ “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

Assign Issue Skill

Assign a GitHub issue to an appropriate owner using repository-defined area ownership and historical assignment patterns.

Inputs

  • An issue reference (owner/repo and issue number, or an issue URL).
  • Optional assignment constraints supplied by the user.

Behavior precedence

Within assignment, explicit current-user instructions take precedence over validated assignment customization, which takes precedence over the static mapping, historical search, confidence, and output defaults below. Neither may change the assignment role, authorize an unrequested write, remove existing assignees, assign unsupported principals, or override global security and repository-scope boundaries.

Assignment Strategies

Static area mapping

Follow the issuelens-config skill and call the issuelens-config tool for domain assignment. The tool uses a configured path when present. When .github/issuelens.yml is absent or omits assignment, it preserves the legacy area-owner search order:

  1. .github/area_owners.md
  2. docs/area_owners.md
  3. area_owners.md

Subject to explicit user constraints, match the issue title, body, labels, and mentioned file paths against the areas' keywords and path patterns. The format is described in references/area_owners_template.md. If the source is built-in, no static mapping exists and historical patterns are the primary strategy. If configuration loading fails, stop and do not assign anyone.

Historical patterns

Use the available GitHub issue-search operation to find similar closed (and recently resolved) issues in the same repository:

  1. Search using the most specific terms from the issue title, body, and labels.
  2. Prefer issues updated in the last 6 to 12 months.
  3. Inspect up to 20 relevant results and count their assignees.
  4. Rank candidates by assignment frequency and similarity to the target issue, and choose the top-ranked individual as the proposed assignee.

Do not infer ownership from issue authorship alone.

By default, this is the primary strategy when the repository has no area-owner file. In that case, always run this search and assign the best-matching individual from the historical data (see the Workflow). Only skip assignment when the search returns no similar issues with a clear, recurring assignee.

Default workflow

  1. Read the target issue and preserve its existing assignees.
  2. Apply the static area mapping when an area-owner file exists.
  3. Analyze similar closed issues for historical assignee patterns.
  4. Combine the evidence:
    • Both strategies agree: high confidence.
    • One strategy produces a clear match: medium confidence.
    • No area-owner file: rely on historical patterns — if similar past issues have a clear, recurring top assignee, treat it as medium confidence and assign that individual.
    • Strategies conflict or evidence is weak: low confidence; report candidates and do not assign automatically.
  5. For a medium- or high-confidence individual user, update the issue with the available assignment operation. Pass the union of existing assignees and the selected username so no current assignment is removed.
  6. Read the issue again and confirm that the selected assignee is present.
  7. Report the decision, evidence, and whether the write was confirmed.

Rules

  • Use only the bundled IssueLens GitHub MCP tools for every GitHub read or write. Pass the explicit owner/repository to every read, search, and write. Never use a shell command, bundled script, direct HTTP request, or GitHub CLI command.
  • Never remove or replace existing assignees.
  • Do not assign a team handle directly; GitHub issue assignees must be individual users. Resolve a team to an eligible individual only when the available GitHub tools provide enough evidence; otherwise report the team as a routing suggestion.
  • Treat a failed or unconfirmed write as a suggestion, not a completed action.
  • If no area-owner file exists, use historical patterns as the primary strategy: search similar past issues and assign the most frequent / most relevant individual assignee. Only skip assignment when the historical evidence yields no reasonable candidate.
  • If no confident match exists, do not guess and do not modify the issue.

Default output

  • Assigned to: the selected username, or None.
  • Confidence: High, Medium, or Low.
  • Reasoning: the area match and historical evidence.
  • Existing assignees: assignees preserved by the update.
  • Action taken: assignment confirmed, assignment failed, or manual review suggested.

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.