Require command outputs and file changes as evidence for code quality claims.
npx skills add https://github.com/TheBushidoCollective/han --skill proof-of-work
Please help me install this Agent Skill. Skill: proof-of-work Source: https://github.com/TheBushidoCollective/han/tree/main/core/skills/proof-of-work Command: npx skills add https://github.com/TheBushidoCollective/han --skill proof-of-work
This Skill enforces evidence-based claims about tests, builds, and quality through verifiable outputs.
After making a claim (e.g., tests pass), provide the exact git status, diffs, and sample outputs as proof.
High-intent search queries and answers about installing and using this skill.
Verifiable proof means showing actual command outputs—git diffs, build logs, test results, and file writes—not just claiming success. Use Write/Edit tools to create or modify files, then run git status, git diff, ls, and cat to document the evidence before reporting completion.
Yes. This approach applies across feature implementation, CI pipelines, code reviews, and agent delegation by requiring concrete outputs such as file writes and build logs as proof that quality checks passed.
Valid evidence includes actual command outputs: git status showing file changes, git diff displaying code modifications, build logs confirming compilation, test result output proving test passage, and file content from cat confirming writes succeeded.
Evidence-based verification creates audit-ready, reproducible records that hold claims accountable. Actual outputs prove quality gates passed rather than relying on assertions, enabling reproducibility and trust in development workflows.
Collect mandatory verification outputs in sequence: perform Write/Edit operations, emit explicit success messages, then capture and report git status, diffs, and file contents as timestamped proof of each verification step completed.
hackefellerBootstrap project templates, install dependencies, configure tooling, and initialize Git.
dsifryRequire fresh verification evidence before reporting completion in coding workflows.
lllllllamaCapture and normalize execution evidence from short non-training commands into structured repro_outputs bundles.
heliohqConvert product changes into end-to-end tests with pass, fail, or skip reports.
