From superpowers-plus
Enumerates all installed skills at runtime, distinguishing auto-triggered from explicit skills. Routes users to the right skill based on their intent.
How this skill is triggered — by the user, by Claude, or both
Slash command
/superpowers-plus:superpowers-helpThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
> **Wrong skill?** Checking skill health → `superpowers-doctor`. Writing new skills → `skill-authoring` / `writing-skills`. Updating skills → `update-superpowers`.
Wrong skill? Checking skill health →
superpowers-doctor. Writing new skills →skill-authoring/writing-skills. Updating skills →update-superpowers.
NEVER hardcode skill names, counts, or routing. All data comes from runtime discovery.
Run this FIRST — every time, no exceptions:
node ~/.codex/superpowers-augment/superpowers-augment.js find-skills
From the output, extract:
Present as a brief summary:
You have [N] skills installed ([X] auto-triggered, [Y] explicit). Auto-triggered skills fire when needed — you don't have to remember them. Tell me what you want to do and I'll match you to the right skill.
If the user describes a task or asks "what should I use for X?", use match-skills:
node ~/.codex/superpowers-augment/superpowers-augment.js match-skills "<user's intent>"
This returns the top 5 matching skills ranked by relevance with scores. Present the top 3 with their descriptions (pulled from the find-skills output, not from memory).
node ~/.codex/superpowers-augment/superpowers-augment.js find-skills superpowers # auto-triggered only
node ~/.codex/superpowers-augment/superpowers-augment.js find-skills explicit # manual only
node ~/.codex/superpowers-augment/superpowers-augment.js use-skill <skill-name>
| Axis | Values | What It Means |
|---|---|---|
| Activation | 🦸 auto-triggered / 🔧 explicit | Auto-triggered skills fire when trigger phrases are detected. Explicit skills must be invoked by name. |
| Source | superpowers: (core) / superpowers-plus: (extended) | Core skills come from Jesse Vincent's obra/superpowers via the bordenet/superpowers fork. Extended skills add domain-specific capabilities. |
The 1% Rule: If there's even a 1% chance a superpower applies, let it fire. Don't suppress with "this is simple."
When showing skills to the user, organize dynamically using the data from find-skills:
match-skills with their answerRun find-skills and present the full output. It's already categorized (auto-triggered vs explicit) and alphabetized.
Run match-skills "<task>" and present the ranked results. The matching engine uses trigger phrases and descriptions to find relevant skills — it knows things this skill file doesn't.
sp-doctor # expands to superpowers-doctor (normal resolution)
spp-doctor # loads from superpowers-plus source repo directly
spo:skill # loads from overlay source repo (requires SP_OVERLAY_SOURCE_DIR)
| Failure | Fix |
|---|---|
| Hardcoding skill names, counts, or routing tables | ALL data must come from find-skills / match-skills at runtime |
| Reporting skills from memory instead of running discovery | Run find-skills before answering — never enumerate from memory |
| Dumping the full catalog as first response | Start with count summary + "what do you want to do?" |
Missing overlay skills from SP_OVERLAY_SOURCE_DIR | Overlay adds skills not in base install — find-skills covers both sources |
| Confusing superpowers vs explicit skills | Two axes: activation (auto/explicit) and source (core/extended) |
| Recommending a skill without confirming it's installed | Run find-skills {name} or match-skills before recommending |
| Stale skill descriptions in output | If find-skills shows > as description, the installed copy needs re-syncing — run install.sh |
| Resource | URL |
|---|---|
| Core superpowers (upstream) | https://github.com/obra/superpowers (Jesse Vincent, MIT) |
| Core superpowers (fork) | https://github.com/bordenet/superpowers |
| superpowers-plus | https://github.com/bordenet/superpowers-plus |
| Architecture | https://github.com/bordenet/superpowers-plus/blob/main/docs/ARCHITECTURE.md |
| Contributing | https://github.com/bordenet/superpowers-plus/blob/main/docs/CONTRIBUTING.md |
npx claudepluginhub bordenet/superpowers-plus --plugin superpowers-plusGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.