omni-pi 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CREDITS.md +28 -0
- package/LICENSE +21 -0
- package/README.md +81 -0
- package/agents/brain.md +24 -0
- package/agents/expert.md +21 -0
- package/agents/planner.md +22 -0
- package/agents/worker.md +21 -0
- package/bin/omni.js +79 -0
- package/extensions/omni-core/index.ts +22 -0
- package/extensions/omni-memory/index.ts +72 -0
- package/extensions/omni-skills/index.ts +11 -0
- package/extensions/omni-status/index.ts +11 -0
- package/package.json +75 -0
- package/prompts/brainstorm.md +15 -0
- package/prompts/spec-template.md +14 -0
- package/prompts/task-template.md +16 -0
- package/skills/omni-escalation/SKILL.md +17 -0
- package/skills/omni-execution/SKILL.md +18 -0
- package/skills/omni-init/SKILL.md +19 -0
- package/skills/omni-planning/SKILL.md +19 -0
- package/skills/omni-verification/SKILL.md +18 -0
- package/src/commands.ts +521 -0
- package/src/config.ts +154 -0
- package/src/context.ts +165 -0
- package/src/contracts.ts +183 -0
- package/src/doctor.ts +225 -0
- package/src/git.ts +135 -0
- package/src/memory.ts +25 -0
- package/src/pi.ts +240 -0
- package/src/planning.ts +303 -0
- package/src/plans.ts +247 -0
- package/src/repo.ts +210 -0
- package/src/skills.ts +308 -0
- package/src/status.ts +105 -0
- package/src/subagents.ts +1031 -0
- package/src/sync.ts +70 -0
- package/src/tasks.ts +141 -0
- package/src/templates.ts +261 -0
- package/src/work.ts +345 -0
- package/src/workflow.ts +375 -0
- package/templates/omni/DECISIONS.md +10 -0
- package/templates/omni/IDEAS.md +13 -0
- package/templates/omni/PROJECT.md +19 -0
- package/templates/omni/SESSION-SUMMARY.md +13 -0
- package/templates/omni/SKILLS.md +21 -0
- package/templates/omni/SPEC.md +11 -0
- package/templates/omni/STATE.md +7 -0
- package/templates/omni/TASKS.md +6 -0
- package/templates/omni/TESTS.md +17 -0
- package/templates/omni/research/README.md +3 -0
- package/templates/omni/specs/README.md +3 -0
- package/templates/omni/tasks/README.md +3 -0
- package/templates/pi/agents/omni-expert.md +13 -0
- package/templates/pi/agents/omni-worker.md +13 -0
|
@@ -0,0 +1,19 @@
|
|
|
1
|
+
# Project
|
|
2
|
+
|
|
3
|
+
## Goal
|
|
4
|
+
|
|
5
|
+
Describe what this project should achieve.
|
|
6
|
+
|
|
7
|
+
## Users
|
|
8
|
+
|
|
9
|
+
- Primary users:
|
|
10
|
+
- Secondary users:
|
|
11
|
+
|
|
12
|
+
## Constraints
|
|
13
|
+
|
|
14
|
+
- Technical constraints:
|
|
15
|
+
- Product constraints:
|
|
16
|
+
|
|
17
|
+
## Success Criteria
|
|
18
|
+
|
|
19
|
+
- What does success look like?
|
|
@@ -0,0 +1,17 @@
|
|
|
1
|
+
# Tests
|
|
2
|
+
|
|
3
|
+
## Project-wide checks
|
|
4
|
+
|
|
5
|
+
-
|
|
6
|
+
|
|
7
|
+
## Task-specific checks
|
|
8
|
+
|
|
9
|
+
-
|
|
10
|
+
|
|
11
|
+
## Retry policy
|
|
12
|
+
|
|
13
|
+
- Worker retries before expert takeover: 2
|
|
14
|
+
|
|
15
|
+
## Escalation threshold
|
|
16
|
+
|
|
17
|
+
- Escalate after repeated failures or when the planner marks the task as high-risk.
|
|
@@ -0,0 +1,13 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: omni-expert
|
|
3
|
+
description: Omni-Pi expert for escalated implementation tasks
|
|
4
|
+
model: anthropic/claude-opus-4-1
|
|
5
|
+
tools: read, grep, find, ls, bash, edit, write
|
|
6
|
+
skill: omni-escalation, omni-verification
|
|
7
|
+
---
|
|
8
|
+
|
|
9
|
+
You are Omni-Pi's expert subagent.
|
|
10
|
+
|
|
11
|
+
Take over difficult or repeatedly failing tasks, fix the root cause, run the required checks when possible, and end with JSON only using this schema:
|
|
12
|
+
|
|
13
|
+
{"summary":"...","verification":{"passed":true,"checksRun":["..."],"failureSummary":[],"retryRecommended":false}}
|
|
@@ -0,0 +1,13 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: omni-worker
|
|
3
|
+
description: Omni-Pi worker for bounded implementation tasks
|
|
4
|
+
model: anthropic/claude-sonnet-4-5
|
|
5
|
+
tools: read, grep, find, ls, bash, edit, write
|
|
6
|
+
skill: omni-execution, omni-verification
|
|
7
|
+
---
|
|
8
|
+
|
|
9
|
+
You are Omni-Pi's worker subagent.
|
|
10
|
+
|
|
11
|
+
Complete the assigned task directly, keep the scope tight, run the required checks when possible, and end with JSON only using this schema:
|
|
12
|
+
|
|
13
|
+
{"summary":"...","verification":{"passed":true,"checksRun":["..."],"failureSummary":[],"retryRecommended":false}}
|