I built agent-thanks to trace which open-source repos one coding task actually used

说明与用途

agent‑thanks 自动识别单个编码任务中使用的开源仓库，生成证据报告，并允许手动点赞，遵守自动化边界。

问题

Maintainers rarely receive task-level signals when their code is used by an automated coding agent, making it hard to acknowledge contributions.

结果

agent-thanks provides a deterministic, rule-based pipeline that identifies exactly which open-source repositories were used during a coding task, generates a human-reviewable report, and enables manual starring with explicit consent.

工作流

• 1. Run agent-thanks against the working tree and a Git baseline (e.g., agent-thanks run --repo . --base HEAD --session path/to/agent-session.log --dry-run).
• 2. The tool compares the tree, detects new dependencies, clones, submodules, and provenance lines, and records evidence for each repo.
• 3. Export the report to Markdown (agent-thanks export .agent-thanks-report.json --output OPEN_SOURCE_USE.md).
• 4. Review the Markdown evidence list; optionally star repositories by running the tool in interactive mode, which prompts for explicit y/N decisions before starring.
• 5. If stars are created, the CLI prints an undo command to revert them if needed.

证据

agent-thanks compares the working tree with a Git baseline and can optionally inspect a local agent transcript.

来源

The evidence pipeline can run unattended. Starring cannot.

来源

If a Star succeeds, the CLI prints an exact Undo command for the Stars created by that invocation.

来源