opencode-session-search

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill indexes and displays past conversation messages, creating a surface where instructions embedded in retrieved data could influence the agent's behavior.
  • Ingestion points: Fetches conversation sessions and messages from the OpenCode API (SKILL.md).
  • Boundary markers: No explicit delimiters or 'ignore' instructions are mentioned to prevent the agent from obeying instructions found within the indexed content.
  • Capability inventory: Performs local command execution (pnpm, pm2) and network operations to local services (SKILL.md).
  • Sanitization: No sanitization or validation of the indexed message content is described in the indexing or search steps.
  • [COMMAND_EXECUTION]: The skill uses the pnpm package manager to execute search and indexing commands and utilizes pm2 for managing the indexer service.
  • [DYNAMIC_EXECUTION]: The command pnpm generate-ecosystem indicates the generation of a PM2 configuration file (ecosystem.config.cjs) at runtime, which is then executed to manage the indexer process.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 10:49 AM
Security Audit — agent-trust-hub — opencode-session-search