~ / rigs / ssztyper / agent-harness

ssztyper/agent-harness

Guardrails, gates and evals for coding agents — enforced in code, not in the prompt. Zero dependencies, Claude Code first.

↗ GitHub ★ 0 MIT updated 1d ago personal setup Claude Code
stack: JavaScript
share on X
ARCHETYPE
Pragmatist
A balanced, no-drama setup: some rules, some tools, nothing extreme.
CONTEXT TAX · EVERY TURN
~1.1k tokens
Featherweight · median rig: 2.3k · breakdown
GUARDRAILS
3/5
Blocks destructive commands · Pre-tool screening hook · No YOLO mode · details
Quality 52/100 Guardrails v2

Copy this rig

# review before running: this installs third-party code
$ npx degit ssztyper/agent-harness/.claude ./rig-agent-harness  # inspect, then merge into .claude/

MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.

$ npx degit ssztyper/agent-harness/skills/decision-record .claude/skills/decision-record
$ npx degit ssztyper/agent-harness/skills/gated-contract .claude/skills/gated-contract

Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.

{
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Bash",
        "hooks": [
          {
            "type": "command",
            "command": "node \"$CLAUDE_PROJECT_DIR\"/hooks/outward-guard.mjs"
          }
        ]
      },
      {
        "matcher": "mcp__.*",
        "hooks": [
          {
            "type": "command",
            "command": "node \"$CLAUDE_PROJECT_DIR\"/hooks/outward-guard.mjs"
          }
        ]
      }
    ]
  }
}

Skills (2)

Hooks (3)

eventmatcherruns
PreToolUseBashnode "$CLAUDE_PROJECT_DIR"/hooks/outward-guard.mjs
PreToolUsemcp__.*node "$CLAUDE_PROJECT_DIR"/hooks/outward-guard.mjs
Stop*node "$CLAUDE_PROJECT_DIR"/hooks/recheck-stop.mjs

Similar rigs

copied ✓