The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process raw text from external files which could contain untrusted instructions.\n- Ingestion points: shaping-session.md reads an 'input pile' which can include transcripts and unstructured prose. SKILL.md requires the agent to 'Re-read the file before each write' to incorporate user edits.\n- Boundary markers: The instructions do not define specific delimiters or guardrails to prevent the agent from following instructions embedded within the processed text.\n- Capability inventory: The skill has the capability to write to local files, specifically appending fragments and drafts to markdown files as specified in fragment-mining.md and shaping-session.md.\n- Sanitization: No sanitization or filtering logic is prescribed for the content read from external files before it is processed or written back to the workspace.

drafting-writing