omh-autopilot

Name: omh-autopilot
Author: witt3rd

117stars9forks

How to install omh-autopilot?

witt3rd/oh-my-hermes/omh-autopilot

$npx -y skills add witt3rd/oh-my-hermes --skill omh-autopilot

Installs into the current project.

Summary

pipeline: interview→plan→execute→QA→verify (idea→code)

Files · 1

View on GitHub

SKILL.md

1	# OMH Autopilot — End-to-End Autonomous Pipeline
2
3	## When to Use
4
5	- End-to-end feature implementation from idea to verified, reviewed code
6	- The user says: "autopilot", "build me", "handle it all", "e2e this"
7
8	## When NOT to Use
9
10	- Single-file changes or trivial tasks (just do them)
11	- You want to stay in one continuous session (autopilot is multi-session)
12	- You only need planning (omh-ralplan) or execution (omh-ralph)
13
14	## Prerequisites
15
16	- The `omh` plugin must be installed (`~/.hermes/plugins/omh/`)
17
18	## Architecture: One Phase Step Per Invocation
19
20	Each autopilot invocation reads state, does ONE unit of work, exits. The caller re-invokes.
21	This preserves fresh context at every level — including during the ralph loop.
22
23	```
24	Invocation 1: Phase 0 — requirements (or skip)
25	Invocation 2: Phase 1 — planning (or skip)
26	Invocations 3-N: Phase 2 — ralph iterations (one per call)
27	Invocation N+1: Phase 3 — QA cycle [FRESH SESSION]
28	Invocation M: Phase 4 — validation round [FRESH SESSION]
29	Final: Phase 5 — cleanup → complete
30	```
31
32	See `references/caller-examples.md` for how to drive the loop.
33
34	## Procedure
35
36	### Step 0: Resolve Instance and Acquire Lock
37
38	Autopilot drives a goal through spec → plan → ralph → QA → validation.
39	Two autopilot sessions on the same goal would race on `autopilot`,
40	`ralph`, and `ralph-tasks` state simultaneously. Use per-instance state
41	+ advisory lock.
42
43	1. Resolve `instance_id` in this order:
44	- If a confirmed spec exists at `.omh/specs/{name}-spec.md`, use
45	`instance_id = "{name}"`.
46	- Else if a plan exists at `.omh/plans/ralplan-{slug}.md`, use
47	`instance_id = "{slug}"`.
48	- Else derive from the goal: `instance_id = kebab(goal)[:60]`.
49	2. Acquire the autopilot lock:
50	```
51	lock = omh_state(action="lock", mode="autopilot",
52	lock_key="{instance_id}",
53	session_id="{HERMES_SESSION_ID or uuid}",
54	holder_note="autopilot driving {goal_or_plan}")
55	```
56	On `acquired=false`, report `held_by`, offer wait/cancel/different
57	goal. Stale-pid auto-release applies.
58	3. Pass `instance_id` to every `omh_state` call in this invocation
59	(autopilot, ralph, ralph-tasks).
60	4. When dispatching to ralph in Phase 2, pass the same
61	`instance_id` in the delegation context so the ralph subagent
62	acquires `mode="ralph"` lock on the same slug.
63	5. Release the autopilot lock at every exit point (paused,
64	blocked, complete, exception):
65	```
66	omh_state(action="unlock", mode="autopilot",
67	lock_key="{instance_id}",
68	session_id="{HERMES_SESSION_ID or uuid}")
69	```
70
71	> Singleton fallback (legacy). Omitting `instance_id` writes
72	> `.omh/state/autopilot-state.json` and skips locking. Acceptable only
73	> when running one autopilot at a time.
74
75	### On Every Invocation: Dispatch
76
77	```
78	state = omh_state(action="read", mode="autopilot", instance_id="{instance_id}")
79	```
80
81	- Not found: Fresh start → Smart Detection (below)
82	- Found: Check `context_checkpoint` flag → if true, clear it and exit (phase boundary)
83	- Check staleness: `state.stale = true` → warn, offer fresh start
84	- Check pause: if `pause_after_phase` matches current completed phase → set phase="paused", exit
85	- Dispatch to current phase handler
86
87	### Smart Detection (Fresh Start)
88
89	When no autopilot state exists, detect artifacts:
90
91	1. Confirmed spec in `.omh/specs/*-spec.md` → create state at Phase 1
92	2. Consensus plan in `.omh/plans/ralplan-*.md` → create state at Phase 2
93	3. Ralph complete (`omh_state(action="check", mode="ralph", instance_id="{instance_id}")` → phase="complete") → create state at Phase 3
94	4. Nothing → create state at Phase 0
95
96	Check for active ralph: `omh_state(action="check", mode="ralph", instance_id="{instance_id}")` → if active, warn about existing session.
97
98	```
99	omh_state(action="write", mode="autopilot", instance_id="{instance_id}", data={
100	"phase": "requirements", "goal": "...", "ralph_iteration": 0,
101	"qa_cycle": 0, "max_qa_cycles": 5, "validation_round": 0,
102	"max_validation_rounds": 3, "validation_verdicts": {},
103	"skip_qa": false, "skip_validation": false, "pause_after_phase": null
104	})
105	```
106
107	### Phase 0: Requirements
108
109	Goal: Ensure a confirmed spec exists.
110
111	1. Check `.omh/specs/*-spec.md` with `status: confirmed` → found? Set `spec_file`, advance to Phase 1, exit
112	2. Not found — assess input:
113	- Concrete (file paths, function names, specific tech): generate inline spec, advance
114	- Vague: Load `omh-deep-interview` and follow it. This phase is interactive.
115	3. Update state: `phase: "planning"`, `spec_file: "<path>"`. Exit.
116
117	For fully autonomous runs: run `omh-deep-interview` separately first.
118
119	### Phase 1: Planning
120
121	Goal:

People who liked this also viewed

6 picks

Type

paperclip-create-agentCreate new agents in Paperclip with governance-aware hiring. Use when you need to inspect adapter configuration options, compare existing agent configs, draft a new agent prompt/config, and submit a hSkillsJul 2026157k5
subagent-driven-developmentUse when executing implementation plans with independent tasks in the current sessionSkillsJul 2026154k259k
dispatching-parallel-agentsUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesSkillsJul 2026143k259k
browser-act-skill-forgeForges reusable Skill packages (SKILL.md + scripts) from website exploration via browser-act — no re-exploration later. Use when: user wants a reusable Skill for any website, needs to understand a sitSkillsJul 202685k4.6k
wizardGenerate an interactive bash wizard that walks a human through a manual procedure — third-party setup, a one-off migration, an A→B state transition — opening URLs, capturing values, confirming each stSkillsJul 202675k182k
wonda-cliUsing the Wonda CLI to generate images, videos, music, and audio from the terminal — plus LinkedIn, Reddit, and X/Twitter research and automationSkillsJul 202666k137

omh-autopilot

117stars9forks

View on GitHub

How to install omh-autopilot?

witt3rd/oh-my-hermes/omh-autopilot

$npx -y skills add witt3rd/oh-my-hermes --skill omh-autopilot

Installs into the current project.

Summary

pipeline: interview→plan→execute→QA→verify (idea→code)

Files · 1

View on GitHub

SKILL.md

1	# OMH Autopilot — End-to-End Autonomous Pipeline
2
3	## When to Use
4
5	- End-to-end feature implementation from idea to verified, reviewed code
6	- The user says: "autopilot", "build me", "handle it all", "e2e this"
7
8	## When NOT to Use
9
10	- Single-file changes or trivial tasks (just do them)
11	- You want to stay in one continuous session (autopilot is multi-session)
12	- You only need planning (omh-ralplan) or execution (omh-ralph)
13
14	## Prerequisites
15
16	- The `omh` plugin must be installed (`~/.hermes/plugins/omh/`)
17
18	## Architecture: One Phase Step Per Invocation
19
20	Each autopilot invocation reads state, does ONE unit of work, exits. The caller re-invokes.
21	This preserves fresh context at every level — including during the ralph loop.
22
23	```
24	Invocation 1: Phase 0 — requirements (or skip)
25	Invocation 2: Phase 1 — planning (or skip)
26	Invocations 3-N: Phase 2 — ralph iterations (one per call)
27	Invocation N+1: Phase 3 — QA cycle [FRESH SESSION]
28	Invocation M: Phase 4 — validation round [FRESH SESSION]
29	Final: Phase 5 — cleanup → complete
30	```
31
32	See `references/caller-examples.md` for how to drive the loop.
33
34	## Procedure
35
36	### Step 0: Resolve Instance and Acquire Lock
37
38	Autopilot drives a goal through spec → plan → ralph → QA → validation.
39	Two autopilot sessions on the same goal would race on `autopilot`,
40	`ralph`, and `ralph-tasks` state simultaneously. Use per-instance state
41	+ advisory lock.
42
43	1. Resolve `instance_id` in this order:
44	- If a confirmed spec exists at `.omh/specs/{name}-spec.md`, use
45	`instance_id = "{name}"`.
46	- Else if a plan exists at `.omh/plans/ralplan-{slug}.md`, use
47	`instance_id = "{slug}"`.
48	- Else derive from the goal: `instance_id = kebab(goal)[:60]`.
49	2. Acquire the autopilot lock:
50	```
51	lock = omh_state(action="lock", mode="autopilot",
52	lock_key="{instance_id}",
53	session_id="{HERMES_SESSION_ID or uuid}",
54	holder_note="autopilot driving {goal_or_plan}")
55	```
56	On `acquired=false`, report `held_by`, offer wait/cancel/different
57	goal. Stale-pid auto-release applies.
58	3. Pass `instance_id` to every `omh_state` call in this invocation
59	(autopilot, ralph, ralph-tasks).
60	4. When dispatching to ralph in Phase 2, pass the same
61	`instance_id` in the delegation context so the ralph subagent
62	acquires `mode="ralph"` lock on the same slug.
63	5. Release the autopilot lock at every exit point (paused,
64	blocked, complete, exception):
65	```
66	omh_state(action="unlock", mode="autopilot",
67	lock_key="{instance_id}",
68	session_id="{HERMES_SESSION_ID or uuid}")
69	```
70
71	> Singleton fallback (legacy). Omitting `instance_id` writes
72	> `.omh/state/autopilot-state.json` and skips locking. Acceptable only
73	> when running one autopilot at a time.
74
75	### On Every Invocation: Dispatch
76
77	```
78	state = omh_state(action="read", mode="autopilot", instance_id="{instance_id}")
79	```
80
81	- Not found: Fresh start → Smart Detection (below)
82	- Found: Check `context_checkpoint` flag → if true, clear it and exit (phase boundary)
83	- Check staleness: `state.stale = true` → warn, offer fresh start
84	- Check pause: if `pause_after_phase` matches current completed phase → set phase="paused", exit
85	- Dispatch to current phase handler
86
87	### Smart Detection (Fresh Start)
88
89	When no autopilot state exists, detect artifacts:
90
91	1. Confirmed spec in `.omh/specs/*-spec.md` → create state at Phase 1
92	2. Consensus plan in `.omh/plans/ralplan-*.md` → create state at Phase 2
93	3. Ralph complete (`omh_state(action="check", mode="ralph", instance_id="{instance_id}")` → phase="complete") → create state at Phase 3
94	4. Nothing → create state at Phase 0
95
96	Check for active ralph: `omh_state(action="check", mode="ralph", instance_id="{instance_id}")` → if active, warn about existing session.
97
98	```
99	omh_state(action="write", mode="autopilot", instance_id="{instance_id}", data={
100	"phase": "requirements", "goal": "...", "ralph_iteration": 0,
101	"qa_cycle": 0, "max_qa_cycles": 5, "validation_round": 0,
102	"max_validation_rounds": 3, "validation_verdicts": {},
103	"skip_qa": false, "skip_validation": false, "pause_after_phase": null
104	})
105	```
106
107	### Phase 0: Requirements
108
109	Goal: Ensure a confirmed spec exists.
110
111	1. Check `.omh/specs/*-spec.md` with `status: confirmed` → found? Set `spec_file`, advance to Phase 1, exit
112	2. Not found — assess input:
113	- Concrete (file paths, function names, specific tech): generate inline spec, advance
114	- Vague: Load `omh-deep-interview` and follow it. This phase is interactive.
115	3. Update state: `phase: "planning"`, `spec_file: "<path>"`. Exit.
116
117	For fully autonomous runs: run `omh-deep-interview` separately first.
118
119	### Phase 1: Planning
120
121	Goal:

People who liked this also viewed

6 picks

Type