migrate-slack-agents

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill fetches updated channel logic and companion skill payloads from the vendor's Git repository (nanocoai/nanoclaw) and executes them using a local driver script (setup/lib/skill-driver.ts) during the migration process.
  • [COMMAND_EXECUTION]: Various CLI operations are performed using pnpm, git, and the project's native ncl tool to inventory existing configurations, create new messaging groups, and update database records.
  • [DATA_EXFILTRATION]: The skill accesses the local environment file (.env) to check for the presence of Slack bot and app tokens. It also reads configuration data and agent identities from the project's SQLite database (data/v2.db). Access is limited to internal verification and migration mapping.
  • [INDIRECT_PROMPT_INJECTION]: The migration process ingests existing configuration and database content to inform the agent's actions, creating a surface where modified database records could influence the migration logic.
  • Ingestion points: Reads agent group metadata, workspace IDs, and wiring configurations from data/v2.db and credential keys from .env.
  • Boundary markers: The instructions specify the use of a sanctioned database wrapper and specific SQL queries to limit the scope of ingested data.
  • Capability inventory: The skill has the ability to execute shell commands, perform network requests (via Git and Slack APIs), and modify the local database.
  • Sanitization: Results from database queries are processed by internal scripts and used as inputs for CLI command generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 02:00 AM
Security Audit — agent-trust-hub — migrate-slack-agents