The Agent Skills Directory

[SAFE]: The skill is a legitimate development tool provided by the vendor 'launchdarkly' for project onboarding.
[COMMAND_EXECUTION]: The skill performs routine development tasks such as installing packages via npm/pip/etc., using npx to manage agent skills, and checking for open ports using lsof. These actions are necessary for its intended functionality.
[EXTERNAL_DOWNLOADS]: Configuration and sub-skills are retrieved from official LaunchDarkly GitHub repositories. These operations use standard tooling and target trusted vendor infrastructure.
[CREDENTIALS_UNSAFE]: While the skill manages sensitive LaunchDarkly SDK keys and API tokens, it implements robust safety protocols:
It enforces the use of environment variables and .env files rather than hardcoding.
It automatically verifies or updates .gitignore to prevent secrets from being committed.
It utilizes mandatory blocking decision points (e.g., D7 in the apply skill and D4-LOCAL in mcp-configure) to ensure the user explicitly chooses how secrets are handled.
It provides clear warnings to the user about the risks associated with AI agents handling sensitive tokens.
[DATA_EXFILTRATION]: The skill includes a 'source' parameter in signup links for attribution purposes. This is a standard marketing practice and does not involve the exfiltration of sensitive user or system data.

onboardingV2