# Skill Selection Evals

> Eval-only skill for measuring skill routing accuracy. Not invoked directly — contains selection evals that test whether the agent picks the correct skill for a given prompt.

- **Type:** Skill
- **Install:** `agentstack add skill-raddue-crucible-skill-selection-evals`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [raddue](https://agentstack.voostack.com/s/raddue)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [raddue](https://github.com/raddue)
- **Source:** https://github.com/raddue/crucible/tree/main/skills/skill-selection-evals

## Install

```sh
agentstack add skill-raddue-crucible-skill-selection-evals
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Skill-Selection Evals

This is not an executable skill. It contains evaluation data for measuring the accuracy of skill selection (routing) decisions.

## Purpose

Crucible's 52 execution evals measure quality once a skill is invoked. Selection evals measure whether the **right skill gets invoked** in the first place.

## Eval Types

- **Direct selection**: Given a prompt, does the agent pick the correct skill?
- **Negative selection**: Given a prompt that sounds like skill X but is not, does the agent avoid the false positive?
- **Context-dependent**: Same verb, different context, different correct skill.
- **Cascade ordering**: Multi-skill tasks requiring correct invocation order.

## Boundaries Tested

1. **test-methodology** — TDD vs test-coverage vs adversarial-tester
2. **review-direction** — temper vs review-feedback
3. **adversarial-scope** — red-team vs inquisitor vs audit vs siege
4. **completion-claims** — verify vs finish
5. **bug-handling** — debugging vs verify vs audit
6. **build-vs-raw-dispatch** — build (full idea→PR pipeline) vs a single-skill dispatch (planning, test-driven-development, …)
7. **ui-implement-vs-verify** — mock-to-unity (implement/fix UI from a mock) vs ui-verify (verify / drift-delta report)
8. **onboarding-entry** — getting-started (silent conversation-start bootstrap) vs workshop (on-demand catalog tour) vs project-init (whole-repo codebase onboarding)
9. **explore-vs-improve** — recon (task-scoped investigation) vs project-init (whole-repo + cross-repo onboarding) vs prospector (architectural-friction discovery)

## Difficulty Ratings

Each eval is rated easy/medium/hard based on routing ambiguity. This enables stratified baseline measurement — distinguishing between improvements that lift hard cases (high value) vs confirming easy cases already work (low signal).

## See Also

- `evals/evals.json` — the eval data
- `GRADING.md` — grading criteria and baseline measurement protocol

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [raddue](https://github.com/raddue)
- **Source:** [raddue/crucible](https://github.com/raddue/crucible)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-raddue-crucible-skill-selection-evals
- Seller: https://agentstack.voostack.com/s/raddue
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.