The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute local shell and PowerShell scripts (scripts/snapshot.sh and scripts/snapshot.ps1) to perform file system operations and project backups.
[DATA_EXFILTRATION]: The snapshot utility programmatically accesses and duplicates sensitive files, explicitly identifying .env files as targets for archival. While the scripts include a verification step to ensure the archive directory is git-ignored, the creation of local copies of credential files increases the risk of data exposure if the archive is accidentally shared or if the environment is compromised.
[INDIRECT_PROMPT_INJECTION]: The skill's workflow depends on reading and interpreting various files from the target repository, which could contain malicious instructions designed to influence the agent's behavior during documentation updates.
Ingestion points: Processes CHANGELOG.md, docs/, AGENTS.md, CLAUDE.md, and CONTRIBUTING.md from the project directory.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to isolate or ignore potentially malicious content within these ingested files.
Capability inventory: The agent has the capability to execute local scripts and perform arbitrary file writes across the repository based on the data it reads.
Sanitization: No validation or sanitization is performed on the content read from repository files before it is used to generate documentation or architectural records.

definition-of-done