Twenty CRM
Analysis
This skill is purpose-aligned for Twenty CRM, but it needs review because it can use a bearer token to create, edit, and delete CRM records without built-in safeguards.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
# If second arg is "destroy", calls the /destroy endpoint. ... curl -sS -X DELETE "$URL" \
The script directly deletes CRM objects, including a destroy mode, using the configured bearer token and without a built-in confirmation, scope check, or recovery guidance.
# Expected file: /Users/jhumanj/clawd/config/twenty.env CONFIG_FILE="/Users/jhumanj/clawd/config/twenty.env"
The script uses a fixed, user-specific absolute config path, while SKILL.md tells users to create `config/twenty.env`; the referenced example file is also not in the manifest.
if [ -f "$CONFIG_FILE" ]; then source "$CONFIG_FILE" fi
The config file is sourced as shell code to load variables; this is common for shell scripts, but any commands placed in that file would execute.
Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.
- `TWENTY_API_KEY` (Bearer token) Scripts load this file automatically.
The skill requires a CRM bearer token even though the registry metadata declares no primary credential or required environment variables; that token is then used by broad REST/GraphQL helpers.
