gaia-research/gaia-skill-tree / .agents/skills/gaia-curate
gaia-curate
VerifiedDiscovery-only Gaia curation for one manually selected source page. Fetches and normalizes up to five real SKILL.md candidates, deduplicates and maps them for human L4 review, then stops without evidence scoring or registry mutation.
| File | Kind | Size |
|---|---|---|
| CURATION-CORE.md | markdown | 12.9 KB |
| LUNA-VIABILITY.md | markdown | 1.2 KB |
| SKILL.md | skill md | 3.3 KB |
| fixtures/generic-snapshot.json | config | 100 B |
| fixtures/luna-viability-expected.json | config | 766 B |
| fixtures/luna-viability-page.json | config | 3.8 KB |
| fixtures/review-ready-packet.json | config | 1.6 KB |
| schemas/discovery-packet-v2.schema.json | config | 11.0 KB |
| schemas/discovery-packet.schema.json | config | 6.9 KB |
| scripts/validate_discovery_packet.py | script | 15.3 KB |
LUNA-VIABILITY.md
View on GitHub# Luna viability receipt
Date: 2026-07-13
A real Hermes Agent one-shot run tested the bounded fixture in `fixtures/luna-viability-page.json` against the separate oracle in `fixtures/luna-viability-expected.json`.
## Runtime
- Harness: Hermes Agent
- Provider: `openai-codex`
- Model: `gpt-5.6-luna`
- Reasoning effort: `minimal`
- API calls: 1
- Tools/research: prohibited by the task contract
## Result
All five candidate-local decisions matched the oracle:
| Candidate | Decision |
|---|---|
| Existing browser implementation | `MAP` to supplied `browser-automation` |
| Exact content duplicate | `DUPLICATE` |
| Malformed non-skill | `NOT_A_SKILL` |
| Ambiguous capability bundle | `DEFER` |
| Copied skill with cited origin | `DUPLICATE` |
Result: **5/5 exact match**. Luna did not invent a generic ID.
## Usage receipt
| Metric | Value |
|---|---:|
| Input tokens | 3,686 |
| Output tokens | 164 |
| Reasoning tokens | 40 |
| Total tokens | 3,850 |
| Estimated cost | $0.00 |
| Cost status | `included` |
The USD field is the Hermes receipt for this subscription-included run, not proof of an external invoice. The test demonstrates viability only for the mechanical five-candidate decision boundary; it does not authorize evidence work or registry mutation.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |