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 nativencltool 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.dband 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