agentleFS
Sign inStart free
ThakurvaibhavThakurvaibhav/platform-agent-harness/skills/upstream-triage/SKILL.md

upstream-triage

Skill2 starsChanged 7 days agoApache-2.0View on GitHub

Decide what to do about a confirmed limitation or bug in a third-party dependency — plugin, controller, operator, library — and draft a clean upstream issue when filing is the answer. Use when you have verified a third-party tool does not do what we need: "is this a known issue", "should we file upstream", "file a bug against X", or after reflex R2 confirms the behavior in the source at your pinned version. Triages wait vs work-around vs file by checking whether the project is maintained and a fix is likely. Drafts and confirms before filing, never files automatically, and never leaks internal names or infrastructure detail into the issue body. Do NOT use for bugs in your own repositories.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before committing to a heavy workaround, knowing whether a fix is imminent changes the calculus.

    line 23
  2. Distinctive:

    If filing: the draft, and after confirmation, the filed URL.

    line 46

≈ “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

Skill: Upstream Triage

When you confirm a limitation or bug in a third-party dependency (plugin, controller, operator, library), decide whether to wait, work around, or file upstream, and draft a clean, internal-info-free issue. Draft-and-confirm before filing; never auto-file.

When to use

  • You've confirmed (ideally by reading the source, Standing Reflex R2) that a third-party tool has a real limitation, bug, or missing feature that costs you work or blocks a goal.
  • Before committing to a heavy workaround, knowing whether a fix is imminent changes the calculus.

When NOT to use

  • The problem is your own config/bug, not the dependency's.
  • Trivial/cosmetic, or the workaround is cheap and the tool is clearly unmaintained.

Procedure

  1. Confirm it's upstream, not you. Reproduce or read the source at the version you run. Rule out misconfiguration.
  2. Search the project -- issues, PRs, and discussions -- for the exact behavior. Capture: is it reported? open/closed? who closed it (maintainer vs. reporter)? any in-flight PR? maintainer intent/roadmap comments. Record issue/PR numbers + URLs + dates.
  3. Assess maintenance health -- latest release + date, release cadence, open vs. closed issue counts, maintainer responsiveness, whether it's the officially-recommended option. Conclude: is a fix likely soon?
  4. Decide:
    • Tracked + fix imminent: link/subscribe, don't file; note the workaround is interim.
    • Untracked (or self-closed/stale) + it costs you: draft an issue.
  5. Draft (generic, redacted). Title + body with: a short summary, a MINIMAL repro using placeholder names (svc-stable/svc-canary, /a, /b -- NO internal service/cluster/org/host names), a link to any prior/related/closed issue, the proposed enhancement, the current workaround, and an offer to send a PR. Strip every internal identifier.
  6. Draft and confirm -- then file. Present the draft + target repo to the user; file only on explicit go: gh issue create --repo <owner/repo> --title "<t>" --body-file <f>. Default is draft-and-confirm; never silently auto-file under the user's identity. Confirm the gh account first. End the body with 🤖 on its own last line — this lands in a stranger's repo under the user's name, so maintainers deserve to know an agent drafted it, and it sets expectations for the follow-up thread. See Agent attribution: 🤖 sign-off in core/protocols/safety-and-handoff.md.
  7. Close the loop. Link the filed issue wherever the limitation is recorded (decision memory, Slack message, PR), and bd remember it.

Output

  • Triage verdict: is it tracked? is a fix likely? recommendation (wait / workaround / file).
  • The strongest issue/PR references (numbers, URLs, dates).
  • If filing: the draft, and after confirmation, the filed URL.

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.

Check your setup, free