The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a platform-native command omh runtime record to track the execution status of the legal review workflow. This command is part of the specific Hermes/OMH ecosystem and is used for internal workflow management rather than executing arbitrary or untrusted code.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and analyze external, untrusted content such as contracts, DPAs, and policies, which presents an attack surface for indirect prompt injection.
Ingestion points: The skill ingests user-provided contract, policy, and regulatory text for analysis as described in SKILL.md and procedure.md.
Boundary markers: The skill includes explicit instructions to 'Distinguish supplied authority from legal interpretation' and 'Fail closed on absent, fabricated, stale, superseded or unverified authority'.
Capability inventory: The skill produces structured artifacts (issue matrices, negotiation prep rows) and invokes the omh CLI tool for runtime logging.
Sanitization: Strict verification procedures are defined in procedure.md, including legal_authority_citation_check and legal_final_determination_guard, which require exact source pinpointing and prohibit final conclusions if authority is unverified.

omh-legal-compliance-review