hermes-labs-copilot

Guided, project-local first experience that helps beginners start making something real.

2 commands · hooks

Maps markdown acceptance items to explicit implementation evidence for advisory spec-drift checks.

1 skill · 2 commands

Runs the released LintLang CLI on demand as a portable audit skill, with a separate automatic post-edit hook for Claude Code.

1 skill · hooks

Runs the released quick-gate JavaScript and TypeScript quality gate and reads its gate-result/v1 verdict.

2 skills

Reports actionable git-state facts before file operations when a checkout is behind its upstream.

hooks

Blocks permanent-delete shell commands and redirects the agent to a recoverable trash workflow.

hooks

Local recovery anchors for Claude Code and Codex sessions, plus evidence-gated evaluation prompts.

1 skill

Receipt-bound completion rail with session-start guidance, advisory cached checks, and boundary enforcement.

hooks

Jailbreak regression benchmark for LLM endpoints with repeatable known-pattern attacks and deterministic scoring.

1 skill

Screens inbound prompts through a local Little Canary service before the host sends them to the main model.

hooks

Runs the deterministic pygate Python quality gate and reads its gate-result/v1 verdict.

1 skill

On-demand static analysis of AI system prompts for contradictions, coverage gaps, priority ambiguities, and meta-paradoxes.

1 skill