The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads and parses content from a local file (.qa-learnings/ledger.md) that may contain untrusted data from previous agent sessions. This data is interpolated into the agent's context and used to generate GitHub issues and pull requests.
Ingestion points: Phase 1 reads the local ledger file into the agent's memory.
Boundary markers: The skill does not implement specific delimiters to isolate the ledger content, relying instead on manual user review in Phase 3.
Capability inventory: The skill can execute shell commands via gh (Phase 4/5) and perform file modifications in temporary cloned repositories (Phase 5).
Sanitization: No specific sanitization or filtering of the ledger content is performed before interpolation into commands or issues.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute GitHub CLI (gh) commands for creating issues, labels, and pull requests. These commands are executed after a user review step in Phase 3.
[EXTERNAL_DOWNLOADS]: In Phase 5, the skill performs a repository fork and clone of neonwatty/qa-skills. These operations are consistent with the skill's stated purpose of contributing to the upstream plugin repository owned by the skill author.

submit-learnings