on-keyday/agent-harness
Parallel Claude Code harness — dispatch tasks to multiple agents across repos/hosts, attach interactively, and let agents message each other over a per-task broker. Toy/dogfood scope.
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
Copy this rig
# review before running: this installs third-party code
$ npx degit on-keyday/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 on-keyday/agent-harness/.agents/skills/codebase-tour .claude/skills/codebase-tour $ npx degit on-keyday/agent-harness/.agents/skills/harness-cli-from-a-tool .claude/skills/harness-cli-from-a-tool $ npx degit on-keyday/agent-harness/.agents/skills/harness-cli .claude/skills/harness-cli $ npx degit on-keyday/agent-harness/.agents/skills/independent-review .claude/skills/independent-review $ npx degit on-keyday/agent-harness/.agents/skills/landing-to-main .claude/skills/landing-to-main $ npx degit on-keyday/agent-harness/.agents/skills/memory-maintenance .claude/skills/memory-maintenance $ npx degit on-keyday/agent-harness/.agents/skills/running-parallel-workers .claude/skills/running-parallel-workers $ npx degit on-keyday/agent-harness/.agents/skills/session-debugging .claude/skills/session-debugging $ npx degit on-keyday/agent-harness/.agents/skills/supervising-workers .claude/skills/supervising-workers $ npx degit on-keyday/agent-harness/.claude/skills/dummy-harness .claude/skills/dummy-harness $ npx degit on-keyday/agent-harness/.claude/skills/implementation-pitfalls .claude/skills/implementation-pitfalls $ npx degit on-keyday/agent-harness/.claude/skills/surface-parity-checklist .claude/skills/surface-parity-checklist
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (12)
codebase-tourharness-cli-from-a-toolharness-cliindependent-reviewlanding-to-mainmemory-maintenancerunning-parallel-workerssession-debuggingsupervising-workersdummy-harnessimplementation-pitfallssurface-parity-checklist
Hooks (1)
| event | matcher | runs |
|---|---|---|
| UserPromptSubmit | * | harness-cli agent inbox --json --user-prompt-submit-hook |
Slash commands (3)
/restart-all/runner-down/runner-up
Permissions
deny (0)
—
ask (0)
—
allow (1)
Bash(harness-cli *)
Similar rigs
cyberpatrolunit/agentready
Generate CLAUDE.md + AGENTS.md tailored to your repo — free, client-side, ~30 frameworks
Minimalist 300 tok · · Q3
sm-wanko/Human-on-exception
Execution harness for AI software engineering: humans own intent and semantic boundaries; agents own execution and verification.
Minimalist 2.0k tok · · Q0
chocomyong/harness-kit
The harness for coding agents - copy-paste operating rules that keep an AI coding agent honest on a codebase that has to survive.
Pragmatist 2.4k tok · · Q23
mitaku/andon
A one-person software factory: AI coding agents take tasks from a ledger through review to merge, and call the human only for decisions only they can make.
Pragmatist 728 tok · · Q0