The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from existing codebases (source code, documentation, and manifests) which could contain malicious instructions targeted at the agent.\n
Ingestion points: Files are read using the read and grep tools during the Discovery phase in references/existing-project.md.\n
Boundary markers: The instructions do not specify the use of delimiters or explicit 'ignore instructions' warnings when processing external file content.\n
Capability inventory: The skill has significant capabilities, including shell access via bash for running builds and tests, as well as write and edit for modifying the filesystem.\n
Sanitization: There is no automated sanitization or filtering of the content ingested from the audited project files, relying instead on manual user verification steps.

app-design