Awesome Hermes PluginsHooks & Observability

lennney/stop-that-shit

⭐ 1959 JavaScript added to this list on 2026-08-24 repository created 2026-08-11

Stop That Shit is a guard that keeps an AI coding agent inside the task the user actually asked for. It pairs a written Skill, which describes a "stop ladder" and a set of task modes, with a machine-executable Guard that enforces those boundaries at the host's hook path instead of trusting the model to restrain itself. The repository ships adapters for four harnesses that share one guard core: Codex, Claude Code, OpenCode and the Hermes Agent CLI. The Hermes side lives in the `.hermes-plugin/` directory, a Hermes plugin manifest (`plugin.yaml`) next to a Python entry point and a bundled CommonJS runtime. Hermes-specific adapter code sits in `src/adapters/hermes-hooks.cjs` and `src/adapters/hermes-tool-classifier.cjs`, which classify each tool call before it runs; `scripts/build-hermes-plugin.cjs` assembles the distributable plugin. Modes are the unit of control. Running in `review` mode declares an intent to read and report, so any write is refused; `change` mode opens a narrower window for edits. The guard checks explicit boundaries on file writes, dependency changes, hashes and subagent budgets, and when a call crosses one it returns a permission deny carrying a reason code, the current state and an event id, for example `MODE_FORBIDS_MUTATION` while armed in review. The agent may still read the repository, but the calls that would change it have to pass the guard. Installation needs Node.js 18 or newer. Each host has its own path: a marketplace entry plus `claude plugin install` for Claude Code, `codex plugin add` for Codex, direct GitHub install for OpenCode 1.18.18 and above, and the plugin directory for Hermes. A local runtime stores cases and evidence, and the repository keeps paired evaluation records for its Codex origins. Users who only want the guidance can install the Skill alone, without the executable guard.

https://github.com/lennney/stop-that-shit

guardhookssafetypermissionstask-boundariescross-harness

Also in Hooks & Observability

rullerzhou-afk/clawd-on-desk

Clawd on Desk is a pixel desktop pet that reacts to your AI coding agent's activity in real time, offering visual feedback for various states and supporting numerous AI agents including Hermes.

8bit64k/cronalytics

Cronalytics is a Hermes Agent plugin providing detailed analytics and observability for scheduled agentic automations, tracking costs, tokens, and performance.

Adolanium/hermes-resetwatch

Hermes Desktop plugin showing live remaining-quota bars and reset timers for Claude, Codex, Cursor, and a dozen other AI provider plans on one page inside the app.

briancaffey/hermes-otel

An OpenTelemetry plugin for Hermes Agent that exports LLM tool calls, model invocations, and API requests as OTel spans to various tracing backends like Phoenix, Langfuse, SigNoz, and Jaeger.

rarf/hermes-quota-plugin

A Hermes Desktop plugin showing live model-provider quota as a status-bar chip and sidebar pane, fed by a local cache that a Python backend refreshes across eight providers.

alexander-akhmetov/grafana-agento11y-hermes

A Hermes Agent plugin recording LLM calls and tool executions as Grafana Agent Observability generations while exporting OpenTelemetry traces and metrics over OTLP.

BULDEE/ai-craftsman-superpowers

Architecture quality gate shipped as a native Hermes plugin, refusing to conclude a coding turn that leaves layer violations or missing strict types and recording every correction as a candidate rule.

nujovich/hermes-telemetry

A comprehensive telemetry plugin for Hermes Agent that provides real-time budget enforcement and detailed observability for AI operations.