From repo-church
Harden assumptions with local evidence and current research before planning. Use for church harden, assumption validation, research hardening, evidence gate, market-confidence checks, stale claim review, or GSD-style pre-plan risk tightening. Drives the `church ledger assumptions` surface and `church bible sources`/`claim-scan`.
How this skill is triggered — by the user, by Claude, or both
Slash command
/repo-church:church-hardenThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
This skill prevents weak assumptions from becoming implementation plans.
This skill prevents weak assumptions from becoming implementation plans.
church context load --root <repo> --format markdown.church ledger init assumptions --root <repo>.church bible sources --root <repo> --follow-local-md --format jsonchurch bible claim-scan --root <repo> --path <repo> --banned <phrases> --required <phrases>verified-local, verified-current, supported-secondary, user-intent, hypothesis, stale, or contradicted.church ledger add assumptions --root <repo> --id A-... --status <status> --summary <claim> --evidence <source> --owner <owner>.church ledger check assumptions --root <repo> before planning.hypothesis, stale, or contradicted high-impact claims to gap closure before planning.## Evidence Gate
Outcome: PASS | PASS_WITH_RISK | HOLD | BLOCK
Reason:
## Assumption Ledger
| ID | Claim | Confidence | Evidence | Impact if wrong | Required validation | Status |
| --- | --- | --- | --- | --- | --- | --- |
## Research Notes
- Source:
- Finding:
- Planning impact:
## Planning Permission
Can proceed: yes|no
Conditions:
## Tooling Used
Sources report:
Claim scan:
Skipped tools and why:
High-impact assumptions must be verified or explicitly deferred with owner, mitigation, and a recheck point. Do not allow implementation planning to rest on unlabeled inference.
npx claudepluginhub chendrizzy/repo-churchCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.