Verify agent-reported actions with git status, ls, and cat output.
npx skills add https://github.com/knopki/dotfiles --skill proof-of-work-knopki
Please help me install this Agent Skill. Skill: proof-of-work Source: https://github.com/knopki/dotfiles/tree/main/home/private_dot_config/opencode/skills/proof-of-work Command: npx skills add https://github.com/knopki/dotfiles --skill proof-of-work-knopki
This Skill ensures that all assertions made about code quality, test results, builds, or verification are backed by concrete, verifiable evidence, preventing blind trust in agent reports and fostering transparency.
git status, ls, cat) as proof of work.git diff to show the exact changes before reporting success.Use the proof-of-work skill to verify any agent's reported completion by showing the output of git status --short.
High-intent search queries and answers about installing and using this skill.
Verifying agent-reported code changes requires enforcing a strict protocol that mandates running `git status`, `ls`, and `cat` to validate file existence and changes before reporting success.
Preventing blind trust in agent reports involves enforcing evidence-based reporting, which requires showing concrete command output like `git diff` to substantiate claims about test results and builds.
Mandatory verification after task completion is enforced by mandating the use of tools like `git status --short` to validate changes and prove the agent's reported actions before any final report is made.
An agent claims a bug is fixed without showing changes due to blind trust in its own reports, which requires a verification protocol mandating `git diff` output to prove the exact changes before reporting success.
This verification protocol works without external dependencies, relying solely on standard system tools like `git`, `ls`, and `cat` to enforce transparency and validate claims about code quality.
Mandating proof of work for agent trust is needed when preventing catastrophic failures from blind trust in agent reports, ensuring all assertions about builds and verification are substantiated with concrete evidence.
1molchuanVerify MCP, Codex, and Gemini endpoints and initialize RPI state.
imbue-aiAnalyze code repositories to identify and document stylistic inconsistencies in markdown reports.
hidonguyenExecute unit, integration, e2e, and UI tests across JavaScript/TypeScript, Python, Go, Rust, and Flutter projects.
yonRun linters and static analysis tools with severity-based output and auto-fix mode.