The Agent Skills Directory

[SAFE]: The skill is a reasoning and documentation tool. It does not contain code, network requests, or attempts to access sensitive system files.
[INDIRECT_PROMPT_INJECTION]: The skill processes external strategy documents as its primary input. This provides an entry point for untrusted data. The workflow mitigates this through evidence-bound gates that require the agent to quote specific passages, anchoring the analysis to the source text.
[COMMAND_EXECUTION]: No shell commands or dynamic context injection patterns were found in the skill files. The instructions focus on structured analysis and file creation in the local working directory.

strategy-review