{"id":51172341,"url":"https://github.com/sean2077/oh-my-agents","last_synced_at":"2026-06-27T01:30:55.762Z","repository":{"id":364981112,"uuid":"1265742696","full_name":"sean2077/oh-my-agents","owner":"sean2077","description":"Lightweight dual-agent CLI + skills toolkit for Claude Code and Codex workflows.","archived":false,"fork":false,"pushed_at":"2026-06-23T06:22:31.000Z","size":789,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-23T07:15:52.226Z","etag":null,"topics":["agent-workflows","ai-agents","claude-code","cli","coding-agents","developer-tools","go","openai-codex","skills","workflow-automation"],"latest_commit_sha":null,"homepage":null,"language":"Go","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/sean2077.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"docs/security-contract.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-06-11T03:32:46.000Z","updated_at":"2026-06-23T06:22:35.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/sean2077/oh-my-agents","commit_stats":null,"previous_names":["sean2077/oh-my-agents"],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/sean2077/oh-my-agents","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sean2077%2Foh-my-agents","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sean2077%2Foh-my-agents/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sean2077%2Foh-my-agents/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sean2077%2Foh-my-agents/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/sean2077","download_url":"https://codeload.github.com/sean2077/oh-my-agents/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sean2077%2Foh-my-agents/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34839004,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-26T02:00:06.560Z","response_time":106,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent-workflows","ai-agents","claude-code","cli","coding-agents","developer-tools","go","openai-codex","skills","workflow-automation"],"created_at":"2026-06-27T01:30:52.392Z","updated_at":"2026-06-27T01:30:55.749Z","avatar_url":"https://github.com/sean2077.png","language":"Go","funding_links":[],"categories":[],"sub_categories":[],"readme":"# oh-my-agents\n\nA single Go binary (`oma`) plus a small set of curated, agent-neutral skills that form a lightweight base layer for everyday AI coding workflows across both Claude Code and Codex.\n\n`oma` solidifies the *mechanical* parts of a workflow (asset install/projection, state, scoring gates, loop stop-judgment, a cross-review pair ledger) into a deterministic, fail-closed CLI. The skills describe only the *non-mechanical* parts: the judgment. They call `oma` commands for everything that should be counted, validated, or persisted.\n\n## Why\n\nThe trigger for this project was a concrete pain point in oh-my-claudecode (OMC): it ships ~40 skills that are **always resident** in the model's context (roughly 15-20k tokens) with **no per-skill disable**. You pay that context tax on every turn whether or not the work needs any of those skills.\n\n`oma` is the opposite bet:\n\n- **You decide what is installed.** Skills are explicit assets you install and remove. Nothing is resident unless you put it there. The four core skills together cost **~275 tokens** of resident surface (name + description), versus OMC's 15-20k, about 2%, and every asset is independently installable.\n- **Mechanical logic belongs in a binary, not a prompt.** Sequence numbers, ambiguity math, threshold gates, stall detection, atomic file writes, and integrity checks are deterministic. They live in `oma`, where they are testable and fail-closed, not re-derived by the model each turn.\n- **Skills stay agent-neutral.** A skill's default path is plain `oma` commands plus markdown, so Claude Code and Codex follow the *same* contract. Host-only accelerations (Claude Code's structured option picker, subagents, plan mode) are clearly-marked optional branches, never the default.\n- **One asset model, two agents.** Assets live in canonical `~/.agents/` and are projected into both `~/.claude/` and `~/.codex/` (symlink on Unix-like hosts; directory junction for skills on native Windows, with managed copy fallback). Install once, available to both. (Hook assets are placed canonically only — oma never writes your host config; you wire hooks into `settings.json`/`hooks.json` by hand, see [`docs/reference/relay-v2-protocol.md`](docs/reference/relay-v2-protocol.md) §12.4.)\n\nThis is CLI + skills, deliberately **not** a Claude Code plugin: a plugin is a Claude-Code-only concept, and the whole point is to stay neutral and lightweight.\n\n## What ships\n\nFour core workflow skills (the \"core4\"):\n\n| Skill | What it does |\n|---|---|\n| **deep-interview** | Socratic requirements crystallization with deterministic ambiguity gating. Turns a vague idea into a `pending approval` spec, never straight to code. |\n| **ralph** | A persistent improvement loop that iterates until a verifier passes — or, under the `score_improvement` keep-policy, until a scored evaluator plateaus. `oma` counts rounds and judges the stop; the agent runs the verifier itself and reports the exit code (and score). |\n| **autopilot** | End-to-end autonomous delivery (clarify -\u003e plan -\u003e implement -\u003e verify -\u003e deliver) with resumable phase state in `oma state`. Pure markdown; no dedicated command group. |\n| **pair-delivery** | Cross-agent delivery over the `oma relay` ledger (plan -\u003e review -\u003e implement -\u003e review -\u003e decision) with an explicit lead and rule-based role-swap escalation. |\n\nOn-demand skills (install only when a task needs them — zero resident cost otherwise):\n\n| Skill | What it does |\n|---|---|\n| **trace** | Adversarial root-cause investigation: competing hypotheses, evidence-strength ranking, self-falsification, ending at the single highest-value discriminating probe. |\n| **analyze** | Read-only deep repository analysis: a ranked, confidence-tagged synthesis with `file:line` evidence and a strict evidence / inference / unknown split. |\n| **best-practice-research** | Bounded external best-practice research with official/upstream sourcing, version/date context, and a terminal read-only handoff. |\n| **research-mission** | Scaffold a falsifiable research/optimization mission (deterministic evaluator contract + candidate ledger) and drive it with ralph's `score_improvement` keep-policy. |\n| **ai-slop-cleaner** | Regression-safe, deletion-first cleanup of AI code slop, gated by a green verifier. |\n| **ultraqa** | Adversarial end-to-end QA as a ralph profile (hostile-scenario matrix). |\n| **skillify** | Capture a repeatable workflow into a new oma skill, gated by a 3-question quality test. |\n\nThe full installed catalog (with lifecycle status) is `oma asset catalog`; `oma asset audit` flags catalog bloat (orphan / oversized / retire) as advisory-only signals.\n\n\u003e **The skills require the `oma` CLI.** They are deliberately *not* standalone prompts: every mechanical step (state, scoring gates, sequence-numbered ledger operations) shells out to `oma`, and a skill invoked without the binary on `PATH` stops at its first command. Install the CLI first, then the skills. By the same principle, skill bodies carry only the core workflow — installation and platform guidance live here in the README, never in a `SKILL.md`.\n\n## Install the CLI\n\nInstall the latest released `oma` into `~/.local/bin`:\n\n```bash\ncurl -fsSL https://raw.githubusercontent.com/sean2077/oh-my-agents/main/scripts/install.sh | bash\n```\n\nThis downloads the prebuilt binary for the **latest GitHub release**, verifies its SHA-256 against the release `checksums.txt`, atomically installs `oma` into `${OMA_INSTALL_BIN_DIR:-$HOME/.local/bin}`, and asserts the installed binary's version — the same fail-closed contract `self-update` uses, no Go toolchain required. It is **fail-closed**: if it cannot resolve a release, match a prebuilt asset, verify the checksum, or confirm the version, it stops with an actionable error — it never silently builds from source or from the unreleased `main` branch.\n\nUseful overrides: `OMA_INSTALL_VERSION=vX.Y.Z` pins a specific release, `OMA_INSTALL_BIN_DIR=/some/bin` changes the destination, and `OMA_INSTALL_FROM_SOURCE=1` opts into a source build (needs `git` + `go`). On Windows, run the same command from Git Bash (it installs `oma.exe`, callable as `oma` once the directory is on `PATH`), or use the native PowerShell installer:\n\n```powershell\nirm https://raw.githubusercontent.com/sean2077/oh-my-agents/main/scripts/install.ps1 | iex\noma version\n```\n\nFor a reproducible, supply-chain-pinned install, fetch the installer **at a release tag** (so the script itself is immutable, not the moving `main`) and pin the version to match:\n\n```bash\nOMA_VERSION=v1.0.0   # a tag from the releases page\ncurl -fsSL \"https://raw.githubusercontent.com/sean2077/oh-my-agents/${OMA_VERSION}/scripts/install.sh\" | OMA_INSTALL_VERSION=\"$OMA_VERSION\" bash\n```\n\nYou can also install from a checkout. This is the preferred self-build path:\nit stamps `oma version` with `git describe`, the short git commit, and\n`-dirty` when the checkout has uncommitted changes.\n\n```bash\ngit clone https://github.com/sean2077/oh-my-agents\ncd oh-my-agents\nmake install\noma version\n```\n\n`make install` uses Go's normal install location (`GOBIN`, or `GOPATH/bin` when\n`GOBIN` is unset), so make sure that directory is on `PATH`. Use `make build`\ninstead when you want a stamped local `./oma` binary in the checkout.\nOn native Windows / Codex Desktop, PowerShell users can skip `make` and run:\n\n```powershell\ngo install -trimpath ./cmd/oma\noma version\n```\n\nEnsure Go's install directory (usually `%USERPROFILE%\\go\\bin`) is on `PATH`.\n\nOnce releases are published, `oma self-update` updates the binary in place from the pinned GitHub Releases (checksum-verified, atomic, with automatic rollback).\n\n## Install the skills\n\n**Prerequisite: the `oma` CLI must be installed and on `PATH`** (see above) — the skills drive it for every mechanical step and do not work without it.\n\n```bash\n# Clean machine: fetch the assets bundle matching your installed oma version\n# (verified against the release checksums.txt; never an unpinned fetch).\noma asset install deep-interview ralph autopilot pair-delivery\n\n# Pin an explicit release, or install from a checkout of this repository:\noma asset install --ref v1.0.0 deep-interview ralph autopilot pair-delivery\n./oma asset install --from assets deep-interview ralph autopilot pair-delivery\n\n# Or directly through the npx skills installer (skills only; not oma-managed)\nnpx skills add sean2077/oh-my-agents -g --agent claude-code codex\n```\n\n## Wire the statusline and hooks (optional)\n\n`oma` never writes your host config. The statusline and the auto-continue\nhooks are **opt-in**: you add them to your own `~/.claude/settings.json` (or\n`~/.codex/hooks.json`) by hand. Use the absolute path to your binary (`which oma`)\nbehind an existence guard so a missing binary degrades silently instead of\nspamming command-not-found.\n\n**Statusline** (a compact one-line view of the core workflow you're currently\nin — relay / ralph / interview / autopilot, each tagged `oma`) — add to\n`~/.claude/settings.json`:\n\n```json\n{\n  \"statusLine\": {\n    \"type\": \"command\",\n    \"command\": \"[ -x '/ABS/PATH/oma' ] || exit 0; exec '/ABS/PATH/oma' statusline\"\n  }\n}\n```\n\n`oma statusline` supersedes the older relay-only `oma relay statusline`. Already\nhave a custom statusline script? Don't replace it — call `oma statusline --json`\nfrom inside it and gate on `.active` (so idle windows stay clean). See\n[`docs/examples/statusline-command.sh`](docs/examples/statusline-command.sh)\nfor a complete working example.\n\n**Auto-continue hooks** (drive the pair-delivery loop without manual nudging) —\nadd to the top-level `hooks` key in `~/.claude/settings.json`. For Codex, the\nStop hook is the main self-continuation path; without a trusted Stop hook it\nfalls back to foreground `oma relay wait`. The dispatcher `oma relay hook\n\u003cevent\u003e` is pure-read and always exits 0, so it can never break your session:\n\n```json\n{\n  \"hooks\": {\n    \"SessionStart\": [{ \"matcher\": \"startup|resume|clear\",\n      \"hooks\": [{ \"type\": \"command\", \"timeout\": 10,\n        \"command\": \"[ -x '/ABS/PATH/oma' ] || exit 0; exec '/ABS/PATH/oma' relay hook SessionStart\" }] }],\n    \"PreToolUse\": [{ \"matcher\": \"^(Edit|Write|MultiEdit)$\",\n      \"hooks\": [{ \"type\": \"command\", \"timeout\": 5,\n        \"command\": \"[ -x '/ABS/PATH/oma' ] || exit 0; exec '/ABS/PATH/oma' relay hook PreToolUse\" }] }],\n    \"Stop\": [{\n      \"hooks\": [{ \"type\": \"command\", \"timeout\": 5,\n        \"command\": \"[ -x '/ABS/PATH/oma' ] || exit 0; exec '/ABS/PATH/oma' relay hook Stop\" }] }]\n  }\n}\n```\n\nFor **Codex**, put the same structure in `~/.codex/hooks.json`, change the\n`PreToolUse` matcher to `^(apply_patch|Edit|Write)$`, then run `/hooks` and\nconfirm the `oma relay hook Stop` entry is trusted. Full field reference\n(matchers, timeouts, guard rationale) is in\n[`docs/reference/relay-v2-protocol.md`](docs/reference/relay-v2-protocol.md) §12.4.\nFor native Windows Codex Desktop, use a PowerShell command string such as:\n\n```json\n{\n  \"hooks\": {\n    \"Stop\": [{\n      \"hooks\": [{ \"type\": \"command\", \"timeout\": 5,\n        \"command\": \"\u0026 'C:\\\\Users\\\\YOU\\\\go\\\\bin\\\\oma.exe' relay hook Stop\" }]\n    }]\n  }\n}\n```\n\nUse the same PowerShell form for `SessionStart` and `PreToolUse`, changing only\nthe final event name.\n\n## Quickstart\n\n```bash\n# See what is installed and healthy\noma asset list --installed\n\n# Check the resident-context budget gate\noma doctor budget --agent claude --profile core4\n```\n\nFor a full walkthrough — install → deep-interview → autopilot → ralph → pair-delivery → failure/resume → upgrade, migrate, and cleanup — see **[docs/tutorial.md](docs/tutorial.md)**.\n\n## Command surface\n\n`oma` is organized into a few command groups (full reference: [`docs/reference/command-tree.md`](docs/reference/command-tree.md)):\n\n- **`oma asset`**: install / list / remove / rollback assets; canonical placement plus per-agent projection. `catalog` derives a status-lifecycle view from manifests; `audit` flags catalog bloat (orphan / oversized / retire), advisory-only.\n- **`oma doctor`**: installation diagnostics and the resident-token budget gate.\n- **`oma interview`**: the solidified surface of deep-interview: scoring math, threshold gate, and state. Math in the CLI; judgment in the agent.\n- **`oma ralph`**: the solidified surface of the loop: round counting, stall detection (`pass_only`) or score-plateau detection (`score_improvement`), terminal-state judgment with a falsifiable receipt. `oma` never runs your verifier.\n- **`oma relay`**: the v2 pair ledger: append-only artifacts, atomic publish with integrity sidecars, sequence reservation, `wait`-based handoff, and a fail-closed approve-close gate (completion receipt binding the reviewed work + a non-lead approve review + its structured review-evidence by content hash).\n- **`oma state`**: generic project-level key/value state for workflows.\n- **`oma config`**, **`oma self-update`**, **`oma version`**.\n\nConventions: `--json` on every query command; `--dry-run` is a global flag that discloses exact paths and writes no persistent target state (remote validation may use auto-cleaned temp files); exit codes are contractual (`0` ok, `1` warn, `2` usage, `3` fail-closed, `4` gate failed; relay `wait` adds `10/11/12`).\n\n## Architecture\n\n```text\n~/.agents/                      canonical asset store (shared with the npx-skills ecosystem)\n  skills/\u003cname\u003e/                skill body\n  agents/\u003cname\u003e.md              subagent\n  hooks/\u003cname\u003e/                 hook (manifest + fragment; canonical-only, wire by hand)\n  prompts/\u003cname\u003e.md             prompt\n        |  projection (symlink on Unix; junction/copy on native Windows; hooks are canonical-only)\n        v\n~/.claude/   ~/.codex/          per-agent directories\n```\n\n- **Assets** are described by a `manifest.json` (`oma-asset/1`) declaring type and target agents. Install places the body in `~/.agents/` and projects it to each target; a registry under `~/.config/oma/` tracks ownership so removal and rollback never touch foreign files.\n- **relay v2** is a fresh protocol (it does not read or write the legacy agent-ledger `.shared/` tree) that keeps the proven principles: append-only ledger, sidecar integrity markers, fail-closed identity, platform-signal authorship. The ledger lives at `\u003crepo\u003e/.oma/relay/`.\n- **Security is fail-closed throughout**: unmanaged-target refusal with backup on `--force`, trusted-root escape checks, POSIX world-writable refusal, duplicate-JSON-key rejection on host configs, mandatory secret scanning before publish, and a checksum-verified self-update trust chain. Details in [`docs/reference/security-contract.md`](docs/reference/security-contract.md).\n\nThe design documents under [`docs/`](docs/) (protocol, command tree, workflows, adapter conformance, schemas, security, config) are the authoritative specification; the implementation follows them rather than the reverse.\n\n## Development\n\n```bash\ngo build ./...\ngo test ./...           # full suite\ngofmt -l .              # must be empty\ngo vet ./...\n```\n\nCI runs the 3-platform test matrix with `-race`, `gofmt`/`vet`/`build`, `golangci-lint`, and `govulncheck` on every push and PR. Releases call the **same** pipeline as a hard gate, then promote the exact built-and-verified artifacts (no rebuild) with a checksums manifest, a strict SemVer tag gate, prerelease/latest classification checks, a build-provenance attestation, and an SBOM. The compatibility contract is [`STABILITY.md`](STABILITY.md).\n\nThis project is itself built through its own `pair-delivery` workflow: every slice is cross-reviewed by a second agent over the `oma relay` ledger before it lands. The skill that describes that process is the one we used to build it.\n\n## Project policies\n\n- **[STABILITY.md](STABILITY.md)** — what is frozen across releases (the compatibility contract).\n- **[SECURITY.md](SECURITY.md)** — supported versions and private vulnerability reporting.\n- **[CONTRIBUTING.md](CONTRIBUTING.md)** — the cross-reviewed delivery process oma is built with.\n\n## License\n\n[MIT](LICENSE) © 2026 sean2077\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsean2077%2Foh-my-agents","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsean2077%2Foh-my-agents","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsean2077%2Foh-my-agents/lists"}