The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local script scripts/claim_ledger.py to validate the project's claim ledger. This is a core part of the skill's documented workflow for research validation.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from a claim ledger file, which serves as a potential surface for indirect prompt injection if the data contains malicious instructions that are processed by the agent or validation scripts.
Ingestion points: 01_story/research/claims.jsonl
Boundary markers: Absent; no specific instructions provided to the agent to ignore embedded commands within the ledger data.
Capability inventory: Local file system read access and command execution through the scripts/claim_ledger.py utility.
Sanitization: No evidence of input validation or sanitization is present in the skill instructions.

evidence-research