The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute git log --oneline -20 to gather recent change history. This is a standard operation for the skill's intended purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted data to populate changelogs and activity logs, which could be used as a vector for indirect prompt injection.
Ingestion points: Extracts data from git commit messages, conversation history, and user input as described in Step 1 of SKILL.md.
Boundary markers: Absent. There are no specific instructions or delimiters used to separate the extracted data from the agent's instructions, nor commands to ignore embedded instructions in the source text.
Capability inventory: The skill utilizes Write, Edit, and Bash tools to modify workspace files and interact with the git repository.
Sanitization: Absent. The instructions do not specify any validation, escaping, or filtering of the extracted content before it is written to markdown files.
[SAFE]: There is a version mismatch between SKILL.md (0.6.4) and plugin.json (1.9.1). This appears to be a minor administrative discrepancy and does not pose a security risk.

atlas-changelog