
Diagnosing Superpowers Sessions
Diagnoses a past or current agent session with cited evidence, then drafts a report or bug issue.
- Repository
- jnMetaCode/superpowers-zh
- Stars
- 8,444
- Licence
- MIT
- Files
- 15 files
What it does
When an AI coding session goes wrong (repeated work, ignored plans, slow or expensive runs, skills that never triggered), this skill makes the agent read the saved session transcript on disk and report what happened, citing exact file and line numbers for each finding. It runs seven parallel analyst sub-agents over different dimensions, writes a structured report, and can search GitHub issues or prepare a redacted bundle for maintainers, but only with the user's approval at each gate.
Great for
- Developers who want to understand why an agent session was slow, costly or off-plan
- Maintainers of the superpowers framework who need a structured, evidence-backed bug report
Skip it if
- Fixing or changing the superpowers skills themselves: the skill is explicitly report-only and leaves triage to humans
- Sharing session data publicly without review: transcripts can contain secrets or personal data, and redaction may miss items, so each file must be checked by the user before sharing
Try asking
- “Why did my last superpowers session take so long? Diagnose it.”
- “Diagnose session abc123 because the agent ignored the plan.”
- “Prepare a redacted bug report bundle for the superpowers maintainers after approval.”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 15 original files, the MIT licence and an attribution note.
diagnosing-superpowers.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip diagnosing-superpowers.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip diagnosing-superpowers.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Name the specific session and the exact problem before starting, since the skill asks clarifying questions first and refuses to analyse without a clear scope.
- Review the redaction log and file list yourself before sharing any bundle, because redaction may be incomplete.
- Ask for a specific dimension such as cost or timing to get a focused answer faster.
What's inside
- prompts/analyst-common.md
- prompts/cost-and-time.md
- prompts/plan-adherence.md
- prompts/quality-evidence.md
- prompts/repeated-work.md
- prompts/request-conflicts.md
- prompts/scrub-audit.md
- prompts/scrub.md
- prompts/similar-session.md
- prompts/skill-timeline.md
- prompts/stumbles.md
- references/context-safety.md
- references/github-issues.md
- references/redaction-policy.md
- SKILL.md
Redistributed unmodified from jnMetaCode/superpowers-zh under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.