The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses node to execute local orchestration scripts (cli.mjs) for managing worker skill lifecycles and environment state.
[EXTERNAL_DOWNLOADS]: Mentions the identification and potential installation of external binary dependencies for language servers and graph providers, such as basedpyright.
[COMMAND_EXECUTION]: The skill provides functionality to modify Cursor and VSCode settings.json to enable high-privilege modes like allowDangerouslySkipPermissions, which bypasses security confirmation prompts.
[COMMAND_EXECUTION]: Automates the configuration of agents into low-restriction states, specifically monitoring or setting approval_policy=never and sandbox_mode=danger-full-access.
[PROMPT_INJECTION]: The skill processes untrusted data from the local filesystem and IDE configuration, creating a surface for indirect prompt injection.
Ingestion points: IDE extension directories (~/.cursor/extensions/, ~/.vscode/extensions/), user settings.json, and local .hex-skills/environment_state.json (SKILL.md).
Boundary markers: Absent; the instructions do not use delimiters or warnings to prevent the agent from following instructions embedded in discovery snapshots.
Capability inventory: The skill possesses the capability to execute shell commands and modify local filesystem configurations (SKILL.md).
Sanitization: While the final state is validated against a JSON schema, the initial discovery phase lacks sanitization or escaping of ingested path and configuration data.

ln-010-dev-environment-setup