skillZs
LIVE SKILL TAGS
>>> LIVE SKILLS INDEX <<<
* OPEN SOURCE *
NO LOGIN, NO TRACKING
REAL INSTALL DATA
← back to all skills
equinor/fusion-skills729 installs

fusion-skill-self-report-bug

Capture Fusion skill workflow failure context and guide a draft-first bug reporting flow with explicit confirmation before any GitHub mutation.

How do I install this agent skill?

npx skills add https://github.com/equinor/fusion-skills --skill fusion-skill-self-report-bug
view source ↗

Is this agent skill safe to install?

  • Gen Agent Trust Hubpass

    The skill is a collection of instructions and templates for drafting bug reports on GitHub following workflow failures. It contains no executable code and includes strong safety guardrails, such as mandatory local drafting and explicit user confirmation for all GitHub actions.

  • Socketpass

    No alerts

  • Snykpass

    Risk: LOW · No issues

  • Runlayerpass

    2/4 files flagged

What does this agent skill do?

Deprecated as of 2026-03-21. This skill has been superseded by the warden agent in fusion-skills, which inlines the full bug reporting workflow and adds proactive frustration detection and skill smell inspection. Use fusion-skills instead.

Install: npx -y skills add equinor/fusion-skills fusion-skills

Self-report Fusion Skill Bugs

When to use

Use this skill when a Fusion skill workflow fails and you need a consistent, triage-ready bug report draft. Treat this as the default failure handoff for any fusion-* skill execution.

Typical triggers:

  • "a fusion-* skill failed"
  • "this skill run failed"
  • "self-report this skill error"
  • "create a bug from this workflow failure"
  • "capture this automation failure for triage"

When not to use

Do not use this skill for:

  • Feature requests or roadmap planning
  • General implementation tasks without observed failure
  • Publishing GitHub changes without explicit user confirmation

Required inputs

Collect before drafting:

  • Target repository for the bug report (default: equinor/fusion-skills)
  • Failing command or workflow step
  • Environment context (OS/shell/runtime/tooling versions if available)
  • Observed output/error evidence
  • Reproduction signal (exact steps or best-effort notes)
  • Optional parent issue number for linking

Instructions

  1. Detect or confirm failure intent.
    • If failure context is missing, ask for command used, expected behavior, actual behavior, and key error output.
  2. Capture failure context.
    • Normalize context into: command, environment, observed error/output, reproducibility, and impact.
  3. Draft locally first.
    • Write .tmp/BUG-skill-failure-<context>.md using assets/issue-templates/skill-workflow-failure-bug.md.
    • Include reproducible steps and explicit observed/expected behavior.
  4. Propose issue metadata.
    • Recommend issue type: Bug.
    • Recommend labels for reliability/automation triage (for example bug, automation, reliability), then validate labels against the target repository before mutation.
    • Ask assignee preference (@me, specific user, or unassigned).
  5. Offer optional relationship linking.
    • If parent issue is provided, prepare to link the new bug as a sub-issue after issue creation.
  6. Ask explicit publish confirmation.
    • Do not run any GitHub mutation until the user explicitly confirms publish.
  7. Mutate only after confirmation.
    • Create issue via MCP issue mutation.
    • Apply labels/assignee.
    • If parent issue was provided, link as sub-issue.
  8. If confirmation is not provided, stop after draft.
    • Return status No GitHub state changes made.

Expected output

Return:

  • Draft bug report file path under .tmp/
  • Captured failure-context summary (command, environment, observed output)
  • Proposed title/body and reproduction steps
  • Proposed issue type + labels + assignee plan
  • Optional parent issue linking plan
  • Explicit status: Awaiting publish confirmation or No GitHub state changes made
  • Created issue URL/number only after confirmed mutation

Safety & constraints

Never:

  • Perform GitHub create/edit/comment/close actions without explicit confirmation
  • Claim failure evidence that was not provided or observed
  • Request or expose secrets/credentials

Always:

  • Keep draft-first behavior
  • Prefer minimal reproducible detail over long narrative
  • Validate label existence in the target repository before applying

Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.

<a href="https://skillzs.dev/skills/equinor/fusion-skills/fusion-skill-self-report-bug">View fusion-skill-self-report-bug on skillZs</a>