crucible
Skill3 starsChanged 29 days agoMITView on GitHub
Write solid unit tests using property-based testing and mutation testing. Use whenever unit tests are being written, fixed, audited, or reviewed in a TypeScript project; whether the user asks directly or test-writing occurs as a step inside another workflow (TDD, feature implementation, bug fixing, code review). Also use when checking test coverage or mutation scores, or when verifying the unit testing configuration. Provides test criteria and processes that other workflows should apply during their test-writing steps.
10 other projects use a version of this.
Adopted by
Most recently taken up by
- luci-efe/skillstheir own version, first seen Nov 1, 2026
- jkitchin/crucibletheir own version, first seen Nov 1, 2026
- IamK77/Skilltheir own version, first seen Oct 12, 2026
- Creeperprol/crucibletheir own version, first seen Sep 19, 2026
- AdityaVasireddy/agent-skillstheir own version, first seen Sep 8, 2026
- raymassie/crucibletheir own version, first seen Sep 7, 2026
Other versions10 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 10
- jkitchin/crucible★ 12same name
- damionrashford/Adversaria★ 9same name
- ryanelian/crucible-agent-skill★ 4same name
- AdityaVasireddy/agent-skills★ 2same name
- IamK77/Skill★ 2same name
- Creeperprol/crucible★ 1same name
- luci-efe/skills★ 0same name
- lukecala/crucible★ 0same name
- raymassie/crucible★ 0same name
- shreyasssk/foundry★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 10 edited versions we read keep these.
8
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 13
When unit tests on a TypeScript project are being written, fixed, or audited — whether requested directly or as a step inside another …
- Distinctive: line 17
When writing end-to-end, integration, UI, or simulation tests.
- Distinctive: line 21
Before activating, check whether the project has opted out of this skill.
- Distinctive: line 26
If any opt-out signal is found, do not activate automatically.
- Distinctive: line 45
Do not modify application code unless specifically prompted to do so.
- Distinctive: line 46
When asked to add tests, do not modify existing tests.
- Distinctive: line 47
When asked to fix tests, do not add new tests.
- Distinctive: line 48
When asked to audit tests, fix existing tests and then add tests as needed to meet test criteria.
“Common” is measured: kept by at least 6 in 10 of the 10 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Crucible
When to use this
When unit tests on a TypeScript project are being written, fixed, or audited — whether requested directly or as a step inside another workflow (e.g., TDD, feature implementation, bug fixing).
When NOT to use this
When writing end-to-end, integration, UI, or simulation tests.
Per-project opt-out
Before activating, check whether the project has opted out of this skill. Look for any of the following:
- A .no-crucible file in the project root
- "crucible: false" anywhere in AGENTS.md or .claude/settings.json
If any opt-out signal is found, do not activate automatically; only apply this skill if the user explicitly requests it in this conversation.
Modes
Pick the mode that matches the task, then read all listed files before starting. Do not work from memory of their contents.
| Mode | When | Read first |
| Write tests | Adding new tests (standalone request or a test-writing step in another workflow) | references/test-criteria.md, references/writing-tests.md, references/assessing-a-module.md, references/mutation-testing.md, references/debrief-template.md |
| Fix tests | Repairing existing tests without adding new ones | references/test-criteria.md, references/assessing-a-module.md, references/mutation-testing.md, references/debrief-template.md |
| Audit tests | Evaluating existing tests, project-wide or scoped | references/test-criteria.md, references/auditing-tests.md, references/assessing-a-module.md, references/mutation-testing.md, references/debrief-template.md |
| Verify setup | Checking that testing dependencies and configs are correct | references/verifying-setup.md |
references/test-criteria.md is mandatory reading in every mode that touches tests: it defines what "done" means.
Rules
- Example commands use
pnpm, but always use the project's package manager. - When invoked standalone:
- Do not modify application code unless specifically prompted to do so.
- When asked to add tests, do not modify existing tests.
- When asked to fix tests, do not add new tests.
- When asked to audit tests, fix existing tests and then add tests as needed to meet test criteria.
- When embedded in another workflow (e.g., a TDD loop that owns code changes): that workflow's rules govern whether application code and in-progress tests may change, but the test criteria in references/test-criteria.md still apply in full, and tests that predate the workflow's current task must not be modified without telling the user. Debrief per the workflow's reporting instructions, following the embedded-workflow rules in references/debrief-template.md.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.