takomi 2.1.13 → 2.1.15

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.
Files changed (39) hide show
  1. package/.pi/agents/architect.md +73 -73
  2. package/.pi/agents/coder.md +70 -70
  3. package/.pi/agents/designer.md +72 -72
  4. package/.pi/agents/orchestrator.md +122 -122
  5. package/.pi/agents/reviewer.md +71 -71
  6. package/.pi/extensions/oauth-router/provider.ts +3 -1
  7. package/.pi/extensions/takomi-context-manager/config.ts +48 -48
  8. package/.pi/extensions/takomi-context-manager/context-router.ts +57 -57
  9. package/.pi/extensions/takomi-context-manager/diagnostics-tools.ts +28 -28
  10. package/.pi/extensions/takomi-context-manager/diagnostics.ts +55 -55
  11. package/.pi/extensions/takomi-context-manager/extension-conflicts.ts +56 -56
  12. package/.pi/extensions/takomi-context-manager/index.ts +56 -56
  13. package/.pi/extensions/takomi-context-manager/model-policy-gate.ts +228 -206
  14. package/.pi/extensions/takomi-context-manager/policy-registry.ts +97 -97
  15. package/.pi/extensions/takomi-context-manager/policy-tools.ts +35 -35
  16. package/.pi/extensions/takomi-context-manager/prerequisite-gates.ts +39 -39
  17. package/.pi/extensions/takomi-context-manager/prompt-rewriter.ts +100 -100
  18. package/.pi/extensions/takomi-context-manager/skill-registry.ts +87 -87
  19. package/.pi/extensions/takomi-context-manager/skill-tools.ts +80 -80
  20. package/.pi/extensions/takomi-context-manager/state.ts +68 -68
  21. package/.pi/extensions/takomi-context-manager/types.ts +77 -77
  22. package/.pi/extensions/takomi-runtime/command-text.ts +10 -2
  23. package/.pi/extensions/takomi-runtime/commands.ts +78 -5
  24. package/.pi/extensions/takomi-runtime/routing-policy.ts +187 -145
  25. package/.pi/extensions/takomi-subagents/native-render.ts +41 -41
  26. package/.pi/extensions/takomi-subagents/pi-subagents-internal.ts +35 -32
  27. package/.pi/extensions/takomi-subagents/run-types.ts +25 -25
  28. package/.pi/prompts/build-prompt.md +259 -259
  29. package/.pi/prompts/design-prompt.md +95 -95
  30. package/.pi/prompts/genesis-prompt.md +140 -140
  31. package/.pi/prompts/prime-prompt.md +110 -110
  32. package/.pi/themes/takomi-aurora.json +88 -88
  33. package/README.md +2 -4
  34. package/assets/.agent/skills/21st-dev-components/SKILL.md +244 -244
  35. package/assets/.agent/skills/anti-gravity/SKILL.md +112 -0
  36. package/assets/.agent/skills/gemini/SKILL.md +14 -223
  37. package/assets/.agent/skills/git-commit-generation/SKILL.md +195 -0
  38. package/package.json +1 -1
  39. package/src/pi-takomi-core/validation.ts +135 -135
@@ -1,110 +1,110 @@
1
- ---
2
- description: Prime the agent with project coding guidelines, styling rules, and folder structure.
3
- ---
4
-
5
- # Workflow: Prime Agent
6
-
7
- > Load the project "brain" — coding guidelines, current work, and verification status.
8
-
9
- ---
10
-
11
- ## Steps
12
-
13
- ### 1. Check Project Health
14
-
15
- Run verification to see current state:
16
-
17
- ```bash
18
- # TypeScript check
19
- npx tsc --noEmit
20
-
21
- # Quick verification (if script exists)
22
- python scripts/vibe-verify.py --quick 2>/dev/null
23
- ```
24
-
25
- **If type-check fails:** There are existing errors. Note them for fixing.
26
-
27
- ---
28
-
29
- ### 2. Load Core Documentation
30
-
31
- ```bash
32
- # Coding Guidelines (The Law)
33
- cat docs/Coding_Guidelines.md 2>/dev/null || cat docs/coding_guidelines.md 2>/dev/null
34
-
35
- # Project Requirements (PRD)
36
- cat docs/Project_Requirements.md 2>/dev/null
37
-
38
- # List mockups
39
- ls docs/mockups/ 2>/dev/null
40
- ```
41
-
42
- ---
43
-
44
- ### 3. Identify Current Work
45
-
46
- ```bash
47
- # List all issues
48
- ls docs/issues/ 2>/dev/null
49
-
50
- # Find incomplete FRs
51
- grep -l "\- \[ \]" docs/issues/*.md 2>/dev/null
52
- ```
53
-
54
- ---
55
-
56
- ### 4. Load nextjs-standards Skill (If Next.js Project)
57
-
58
- If `next.config.*` or `package.json` contains `"next"`:
59
-
60
- ```bash
61
- # Read the skill
62
- cat .agent/skills/nextjs-standards/SKILL.md 2>/dev/null
63
- ```
64
-
65
- ---
66
-
67
- ### 5. State Context Aloud
68
-
69
- After loading, acknowledge:
70
-
71
- "✅ **Agent Primed.**
72
-
73
- **Project Health:**
74
- - TypeScript: [PASS/FAIL - X errors]
75
- - Lint: [PASS/FAIL]
76
-
77
- **Context Loaded:**
78
- - Coding Guidelines: [Found/Not Found]
79
- - Mockups: [X files found / None]
80
-
81
- **Current Work:**
82
- - Incomplete FRs: FR-XXX, FR-YYY
83
- - Next up: FR-XXX: [Title]
84
-
85
- **Rules I will follow:**
86
- - `tsc --noEmit` after every file edit
87
- - Reference issue file for each FR
88
- - Mark acceptance criteria as I complete them
89
-
90
- What would you like me to work on?"
91
-
92
- ---
93
-
94
- ## Quick Reference
95
-
96
- | Command | Purpose |
97
- |---------|---------|
98
- | `npx tsc --noEmit` | Check for type errors |
99
- | `python scripts/vibe-verify.py --quick` | Quick verification |
100
- | `cat docs/Coding_Guidelines.md` | Load the law |
101
- | `ls docs/issues/` | See all FRs |
102
-
103
- ---
104
-
105
- ## When to Use
106
-
107
- - **Start of session** — Before any work
108
- - **After `/agent_reset`** — Reload context after reset
109
- - **Context lost** — Agent seems confused
110
- - **Before complex work** — Ensure full context
1
+ ---
2
+ description: Prime the agent with project coding guidelines, styling rules, and folder structure.
3
+ ---
4
+
5
+ # Workflow: Prime Agent
6
+
7
+ > Load the project "brain" — coding guidelines, current work, and verification status.
8
+
9
+ ---
10
+
11
+ ## Steps
12
+
13
+ ### 1. Check Project Health
14
+
15
+ Run verification to see current state:
16
+
17
+ ```bash
18
+ # TypeScript check
19
+ npx tsc --noEmit
20
+
21
+ # Quick verification (if script exists)
22
+ python scripts/vibe-verify.py --quick 2>/dev/null
23
+ ```
24
+
25
+ **If type-check fails:** There are existing errors. Note them for fixing.
26
+
27
+ ---
28
+
29
+ ### 2. Load Core Documentation
30
+
31
+ ```bash
32
+ # Coding Guidelines (The Law)
33
+ cat docs/Coding_Guidelines.md 2>/dev/null || cat docs/coding_guidelines.md 2>/dev/null
34
+
35
+ # Project Requirements (PRD)
36
+ cat docs/Project_Requirements.md 2>/dev/null
37
+
38
+ # List mockups
39
+ ls docs/mockups/ 2>/dev/null
40
+ ```
41
+
42
+ ---
43
+
44
+ ### 3. Identify Current Work
45
+
46
+ ```bash
47
+ # List all issues
48
+ ls docs/issues/ 2>/dev/null
49
+
50
+ # Find incomplete FRs
51
+ grep -l "\- \[ \]" docs/issues/*.md 2>/dev/null
52
+ ```
53
+
54
+ ---
55
+
56
+ ### 4. Load nextjs-standards Skill (If Next.js Project)
57
+
58
+ If `next.config.*` or `package.json` contains `"next"`:
59
+
60
+ ```bash
61
+ # Read the skill
62
+ cat .agent/skills/nextjs-standards/SKILL.md 2>/dev/null
63
+ ```
64
+
65
+ ---
66
+
67
+ ### 5. State Context Aloud
68
+
69
+ After loading, acknowledge:
70
+
71
+ "✅ **Agent Primed.**
72
+
73
+ **Project Health:**
74
+ - TypeScript: [PASS/FAIL - X errors]
75
+ - Lint: [PASS/FAIL]
76
+
77
+ **Context Loaded:**
78
+ - Coding Guidelines: [Found/Not Found]
79
+ - Mockups: [X files found / None]
80
+
81
+ **Current Work:**
82
+ - Incomplete FRs: FR-XXX, FR-YYY
83
+ - Next up: FR-XXX: [Title]
84
+
85
+ **Rules I will follow:**
86
+ - `tsc --noEmit` after every file edit
87
+ - Reference issue file for each FR
88
+ - Mark acceptance criteria as I complete them
89
+
90
+ What would you like me to work on?"
91
+
92
+ ---
93
+
94
+ ## Quick Reference
95
+
96
+ | Command | Purpose |
97
+ |---------|---------|
98
+ | `npx tsc --noEmit` | Check for type errors |
99
+ | `python scripts/vibe-verify.py --quick` | Quick verification |
100
+ | `cat docs/Coding_Guidelines.md` | Load the law |
101
+ | `ls docs/issues/` | See all FRs |
102
+
103
+ ---
104
+
105
+ ## When to Use
106
+
107
+ - **Start of session** — Before any work
108
+ - **After `/agent_reset`** — Reload context after reset
109
+ - **Context lost** — Agent seems confused
110
+ - **Before complex work** — Ensure full context
@@ -1,88 +1,88 @@
1
- {
2
- "$schema": "https://raw.githubusercontent.com/badlogic/pi-mono/main/packages/coding-agent/src/modes/interactive/theme/theme-schema.json",
3
- "name": "takomi-aurora",
4
- "vars": {
5
- "void": "#070A13",
6
- "obsidian": "#0B1020",
7
- "deep": "#11162A",
8
- "glass": "#18213A",
9
- "glassHot": "#251633",
10
- "glassCool": "#0D2430",
11
- "line": "#2B3558",
12
- "lineGlow": "#3D4B7A",
13
- "snow": "#EAF2FF",
14
- "mist": "#AAB8D6",
15
- "haze": "#6F7EA3",
16
- "neonCyan": "#2EF3FF",
17
- "electricBlue": "#4D8DFF",
18
- "plasmaViolet": "#B56CFF",
19
- "hotMagenta": "#FF4FD8",
20
- "solarGold": "#FFD166",
21
- "acidGreen": "#6DFFB8",
22
- "ember": "#FF6B6B"
23
- },
24
- "colors": {
25
- "accent": "neonCyan",
26
- "border": "line",
27
- "borderAccent": "hotMagenta",
28
- "borderMuted": "lineGlow",
29
- "success": "acidGreen",
30
- "error": "ember",
31
- "warning": "solarGold",
32
- "muted": "mist",
33
- "dim": "haze",
34
- "text": "snow",
35
- "thinkingText": "mist",
36
-
37
- "selectedBg": "glassHot",
38
- "userMessageBg": "deep",
39
- "userMessageText": "snow",
40
- "customMessageBg": "glassCool",
41
- "customMessageText": "snow",
42
- "customMessageLabel": "hotMagenta",
43
-
44
- "toolPendingBg": "glassCool",
45
- "toolSuccessBg": "#0D2C24",
46
- "toolErrorBg": "#321622",
47
- "toolTitle": "neonCyan",
48
- "toolOutput": "snow",
49
-
50
- "mdHeading": "hotMagenta",
51
- "mdLink": "neonCyan",
52
- "mdLinkUrl": "electricBlue",
53
- "mdCode": "solarGold",
54
- "mdCodeBlock": "snow",
55
- "mdCodeBlockBorder": "plasmaViolet",
56
- "mdQuote": "mist",
57
- "mdQuoteBorder": "hotMagenta",
58
- "mdHr": "lineGlow",
59
- "mdListBullet": "neonCyan",
60
-
61
- "toolDiffAdded": "acidGreen",
62
- "toolDiffRemoved": "ember",
63
- "toolDiffContext": "mist",
64
-
65
- "syntaxComment": "haze",
66
- "syntaxKeyword": "hotMagenta",
67
- "syntaxFunction": "neonCyan",
68
- "syntaxVariable": "snow",
69
- "syntaxString": "acidGreen",
70
- "syntaxNumber": "solarGold",
71
- "syntaxType": "electricBlue",
72
- "syntaxOperator": "plasmaViolet",
73
- "syntaxPunctuation": "mist",
74
-
75
- "thinkingOff": "lineGlow",
76
- "thinkingMinimal": "plasmaViolet",
77
- "thinkingLow": "electricBlue",
78
- "thinkingMedium": "neonCyan",
79
- "thinkingHigh": "hotMagenta",
80
- "thinkingXhigh": "ember",
81
- "bashMode": "solarGold"
82
- },
83
- "export": {
84
- "pageBg": "#060813",
85
- "cardBg": "#0E1426",
86
- "infoBg": "#1A1730"
87
- }
88
- }
1
+ {
2
+ "$schema": "https://raw.githubusercontent.com/badlogic/pi-mono/main/packages/coding-agent/src/modes/interactive/theme/theme-schema.json",
3
+ "name": "takomi-aurora",
4
+ "vars": {
5
+ "void": "#070A13",
6
+ "obsidian": "#0B1020",
7
+ "deep": "#11162A",
8
+ "glass": "#18213A",
9
+ "glassHot": "#251633",
10
+ "glassCool": "#0D2430",
11
+ "line": "#2B3558",
12
+ "lineGlow": "#3D4B7A",
13
+ "snow": "#EAF2FF",
14
+ "mist": "#AAB8D6",
15
+ "haze": "#6F7EA3",
16
+ "neonCyan": "#2EF3FF",
17
+ "electricBlue": "#4D8DFF",
18
+ "plasmaViolet": "#B56CFF",
19
+ "hotMagenta": "#FF4FD8",
20
+ "solarGold": "#FFD166",
21
+ "acidGreen": "#6DFFB8",
22
+ "ember": "#FF6B6B"
23
+ },
24
+ "colors": {
25
+ "accent": "neonCyan",
26
+ "border": "line",
27
+ "borderAccent": "hotMagenta",
28
+ "borderMuted": "lineGlow",
29
+ "success": "acidGreen",
30
+ "error": "ember",
31
+ "warning": "solarGold",
32
+ "muted": "mist",
33
+ "dim": "haze",
34
+ "text": "snow",
35
+ "thinkingText": "mist",
36
+
37
+ "selectedBg": "glassHot",
38
+ "userMessageBg": "deep",
39
+ "userMessageText": "snow",
40
+ "customMessageBg": "glassCool",
41
+ "customMessageText": "snow",
42
+ "customMessageLabel": "hotMagenta",
43
+
44
+ "toolPendingBg": "glassCool",
45
+ "toolSuccessBg": "#0D2C24",
46
+ "toolErrorBg": "#321622",
47
+ "toolTitle": "neonCyan",
48
+ "toolOutput": "snow",
49
+
50
+ "mdHeading": "hotMagenta",
51
+ "mdLink": "neonCyan",
52
+ "mdLinkUrl": "electricBlue",
53
+ "mdCode": "solarGold",
54
+ "mdCodeBlock": "snow",
55
+ "mdCodeBlockBorder": "plasmaViolet",
56
+ "mdQuote": "mist",
57
+ "mdQuoteBorder": "hotMagenta",
58
+ "mdHr": "lineGlow",
59
+ "mdListBullet": "neonCyan",
60
+
61
+ "toolDiffAdded": "acidGreen",
62
+ "toolDiffRemoved": "ember",
63
+ "toolDiffContext": "mist",
64
+
65
+ "syntaxComment": "haze",
66
+ "syntaxKeyword": "hotMagenta",
67
+ "syntaxFunction": "neonCyan",
68
+ "syntaxVariable": "snow",
69
+ "syntaxString": "acidGreen",
70
+ "syntaxNumber": "solarGold",
71
+ "syntaxType": "electricBlue",
72
+ "syntaxOperator": "plasmaViolet",
73
+ "syntaxPunctuation": "mist",
74
+
75
+ "thinkingOff": "lineGlow",
76
+ "thinkingMinimal": "plasmaViolet",
77
+ "thinkingLow": "electricBlue",
78
+ "thinkingMedium": "neonCyan",
79
+ "thinkingHigh": "hotMagenta",
80
+ "thinkingXhigh": "ember",
81
+ "bashMode": "solarGold"
82
+ },
83
+ "export": {
84
+ "pageBg": "#060813",
85
+ "cardBg": "#0E1426",
86
+ "infoBg": "#1A1730"
87
+ }
88
+ }
package/README.md CHANGED
@@ -41,10 +41,6 @@ takomi install all
41
41
  takomi init
42
42
  ```
43
43
 
44
- ### Subagent execution credit
45
-
46
- Takomi's Pi-native subagent execution and terminal UI build on **[`pi-subagents`](https://github.com/nicobailon/pi-subagents)** by **Nico Bailon**. That package provides the underlying Pi extension for delegated subagent runs, including the native subagent result renderer, live progress/status display, single/parallel/chain execution support, session/artifact handling, and related subagent tooling. Takomi adds its own lifecycle orchestration, model-routing policy, workflow metadata, board/checklist context, and agent conventions on top of that foundation.
47
-
48
44
  ### Context Manager
49
45
 
50
46
  Takomi now ships a Pi-native `takomi-context-manager` extension. It reduces prompt bloat with progressive context loading:
@@ -480,6 +476,8 @@ Externally sourced skills in this bundle retain credit to their upstream creator
480
476
  - **Gemini CLI**: Custom VibeCode skill for large-context processing with Gemini 3 Pro.
481
477
  - **Google Stitch Skills**: From [google-labs-code/stitch-skills](https://github.com/google-labs-code/stitch-skills) — Design-to-code suite including **design-md**, **enhance-prompt**, **stitch-loop**, **react-components**, and **shadcn-ui**.
482
478
  - **Jules**: From [sanjay3290/ai-skills](https://github.com/sanjay3290/ai-skills) — delegate coding tasks to Google Jules AI agent.
479
+ - **Subagent Execution**: Built on **[`pi-subagents`](https://github.com/nicobailon/pi-subagents)** by **Nico Bailon** — providing the underlying Pi extension for delegated subagent runs (result rendering, live progress, single/parallel/chain execution, session/artifact handling, and related subagent tooling), upon which Takomi adds its own lifecycle orchestration, model-routing policy, and workflow metadata.
480
+ - **Git Commit Generation**: From the **[`kilocode`](https://github.com/Kilo-Org/kilocode)** repository by **Kilo-Org** (specifically, [git-commit-generation.md](https://github.com/Kilo-Org/kilocode/blob/main/packages/kilo-docs/pages/code-with-ai/features/git-commit-generation.md)) — enabling automated, high-quality conventional git commit messages based on staged changes.
483
481
 
484
482
  ## 📄 License
485
483