vibesuite 1.3.3 → 2.0.2
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/README.md +75 -6
- package/assets/.agent/skills/avoid-feature-creep/SKILL.md +307 -0
- package/assets/.agent/skills/avoid-feature-creep/agents/openai.yaml +3 -0
- package/assets/.agent/skills/avoid-feature-creep/assets/large-logo.png +0 -0
- package/assets/.agent/skills/avoid-feature-creep/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex/SKILL.md +62 -0
- package/assets/.agent/skills/convex/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-agents/SKILL.md +516 -0
- package/assets/.agent/skills/convex-agents/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-agents/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-agents/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-best-practices/SKILL.md +369 -0
- package/assets/.agent/skills/convex-best-practices/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-best-practices/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-best-practices/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-component-authoring/SKILL.md +457 -0
- package/assets/.agent/skills/convex-component-authoring/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-component-authoring/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-component-authoring/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-cron-jobs/SKILL.md +604 -0
- package/assets/.agent/skills/convex-cron-jobs/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-cron-jobs/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-cron-jobs/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-file-storage/SKILL.md +467 -0
- package/assets/.agent/skills/convex-file-storage/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-file-storage/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-file-storage/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-functions/SKILL.md +458 -0
- package/assets/.agent/skills/convex-functions/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-functions/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-functions/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-http-actions/SKILL.md +733 -0
- package/assets/.agent/skills/convex-http-actions/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-http-actions/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-http-actions/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-migrations/SKILL.md +712 -0
- package/assets/.agent/skills/convex-migrations/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-migrations/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-migrations/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-realtime/SKILL.md +443 -0
- package/assets/.agent/skills/convex-realtime/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-realtime/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-realtime/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-schema-validator/SKILL.md +400 -0
- package/assets/.agent/skills/convex-schema-validator/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-schema-validator/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-schema-validator/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-security-audit/SKILL.md +539 -0
- package/assets/.agent/skills/convex-security-audit/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-security-audit/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-security-audit/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/convex-security-check/SKILL.md +378 -0
- package/assets/.agent/skills/convex-security-check/agents/openai.yaml +3 -0
- package/assets/.agent/skills/convex-security-check/assets/large-logo.png +0 -0
- package/assets/.agent/skills/convex-security-check/assets/small-logo.svg +17 -0
- package/assets/.agent/skills/github-ops/SKILL.md +4 -4
- package/assets/.agent/skills/google-trends/SKILL.md +7 -7
- package/assets/.agent/skills/optimize-agent-context/SKILL.md +97 -0
- package/assets/.agent/skills/youtube-pipeline/SKILL.md +10 -10
- package/assets/.agent/workflows/LEGACY/init_smart_ops.md +2 -2
- package/assets/.agent/workflows/agent_reset.md +4 -6
- package/assets/.agent/workflows/mode-orchestrator.md +17 -22
- package/assets/.agent/workflows/mode-visionary.md +3 -10
- package/assets/.agent/workflows/optimize-agent-context.md +54 -0
- package/assets/.agent/workflows/remotion-build.md +17 -17
- package/assets/.agent/workflows/stitch.md +4 -4
- package/assets/VibeCode-Agents/vibe-orchestrator.yaml +14 -33
- package/assets/VibeCode-Agents/vibe-visionary.yaml +3 -13
- package/package.json +1 -1
- package/src/cli.js +416 -20
- package/src/harness.js +281 -0
- package/src/store.js +239 -0
- package/assets/VibeCode-Agents/custom_modes.yaml +0 -979
|
@@ -52,26 +52,18 @@ Use `/mode-orchestrator` when:
|
|
|
52
52
|
|
|
53
53
|
## Phase 1: Ecosystem Scan (MANDATORY)
|
|
54
54
|
|
|
55
|
-
### Scan Skills
|
|
55
|
+
### Scan Skills & Workflows
|
|
56
56
|
|
|
57
|
+
**Your system prompt lists all available skills and workflows with their full absolute paths.** Check there FIRST — do NOT guess paths.
|
|
58
|
+
|
|
59
|
+
If your system prompt doesn't list them, fall back to scanning the filesystem:
|
|
57
60
|
```bash
|
|
58
|
-
# MANDATORY: Do not skip this
|
|
59
|
-
# Project-local skills
|
|
60
61
|
ls .agent/skills/ 2>/dev/null
|
|
61
|
-
|
|
62
|
-
|
|
63
|
-
ls ~/.gemini/antigravity/skills/ 2>/dev/null # Gemini / Antigravity
|
|
64
|
-
ls ~/.kilocode/skills/ 2>/dev/null # KiloCode
|
|
62
|
+
ls .agent/workflows/ 2>/dev/null
|
|
63
|
+
ls .agent/global_workflows/ 2>/dev/null
|
|
65
64
|
```
|
|
66
65
|
|
|
67
66
|
Create a **Skills Registry** — which skills are relevant and which tasks they apply to.
|
|
68
|
-
|
|
69
|
-
### Scan Workflows
|
|
70
|
-
|
|
71
|
-
```bash
|
|
72
|
-
ls .agent/workflows/
|
|
73
|
-
```
|
|
74
|
-
|
|
75
67
|
Create a **Workflows Registry** — which workflows map to which task phases.
|
|
76
68
|
|
|
77
69
|
---
|
|
@@ -122,20 +114,23 @@ Each task file (`01_task_name.task.md`) includes:
|
|
|
122
114
|
## 🔧 Agent Setup (DO THIS FIRST)
|
|
123
115
|
|
|
124
116
|
### Workflow to Follow
|
|
125
|
-
>
|
|
117
|
+
> Read the `/vibe-build` workflow. Your system prompt lists all workflows with
|
|
118
|
+
> their full paths — use `view_file` on the path listed there.
|
|
119
|
+
> Fallback: `cat .agent/workflows/vibe-build.md`
|
|
120
|
+
> Fallback: `cat .agent/global_workflows/vibe-build.md`
|
|
126
121
|
|
|
127
122
|
### Prime Agent Context
|
|
128
123
|
> MANDATORY: Run `/vibe-primeAgent` first
|
|
129
124
|
|
|
130
125
|
### Required Skills
|
|
131
|
-
>
|
|
132
|
-
>
|
|
133
|
-
> | nextjs-standards | `[skills-dir]/nextjs-standards/SKILL.md` | Next.js project |
|
|
126
|
+
> **Your system prompt lists all skills with their absolute paths.**
|
|
127
|
+
> Look up each skill below in your system prompt and `view_file` its SKILL.md.
|
|
134
128
|
>
|
|
135
|
-
>
|
|
136
|
-
|
|
137
|
-
|
|
138
|
-
>
|
|
129
|
+
> | Skill | Why |
|
|
130
|
+
> |-------|-----|
|
|
131
|
+
> | nextjs-standards | Next.js project |
|
|
132
|
+
>
|
|
133
|
+
> Fallback (if not in system prompt): `ls .agent/skills/`
|
|
139
134
|
```
|
|
140
135
|
|
|
141
136
|
Then: Objective, Scope, Context, Definition of Done, Expected Artifacts, Constraints.
|
|
@@ -78,21 +78,14 @@ Batch ALL questions into ONE message. Skip anything already answered:
|
|
|
78
78
|
|
|
79
79
|
### 1c. Scan Skills & Workflows
|
|
80
80
|
|
|
81
|
+
**Your system prompt lists all available skills and workflows with their full absolute paths.** Check there FIRST.
|
|
82
|
+
|
|
83
|
+
Fallback (if not in system prompt):
|
|
81
84
|
```bash
|
|
82
|
-
# MANDATORY: Scan before making decisions
|
|
83
|
-
# Project-local skills
|
|
84
85
|
ls .agent/skills/ 2>/dev/null
|
|
85
|
-
|
|
86
|
-
# Global skills (path varies by AI client)
|
|
87
|
-
ls ~/.gemini/antigravity/skills/ 2>/dev/null # Gemini / Antigravity
|
|
88
|
-
ls ~/.kilocode/skills/ 2>/dev/null # KiloCode
|
|
89
|
-
|
|
90
|
-
# Workflows (always project-local)
|
|
91
86
|
ls .agent/workflows/ 2>/dev/null
|
|
92
87
|
```
|
|
93
88
|
|
|
94
|
-
Use whichever location exists. The agent's system prompt may also list available skills.
|
|
95
|
-
|
|
96
89
|
Note which skills and workflows are relevant — these will be injected into the Orchestrator handoff.
|
|
97
90
|
|
|
98
91
|
---
|
|
@@ -0,0 +1,54 @@
|
|
|
1
|
+
---
|
|
2
|
+
description: Create, audit, or optimize agent.md / claude.md / cursorrules files using the Band-Aid Philosophy. Strips bloat, enforces minimalism.
|
|
3
|
+
---
|
|
4
|
+
|
|
5
|
+
# Workflow: Optimize Agent Context
|
|
6
|
+
|
|
7
|
+
> Shortcut to the `optimize-agent-context` skill. Creates minimal, effective AI agent instruction files.
|
|
8
|
+
|
|
9
|
+
---
|
|
10
|
+
|
|
11
|
+
## Steps
|
|
12
|
+
|
|
13
|
+
### 1. Load the Skill
|
|
14
|
+
|
|
15
|
+
```bash
|
|
16
|
+
cat .agent/skills/optimize-agent-context/SKILL.md
|
|
17
|
+
```
|
|
18
|
+
|
|
19
|
+
Read the full skill instructions before proceeding.
|
|
20
|
+
|
|
21
|
+
---
|
|
22
|
+
|
|
23
|
+
### 2. Determine Mode
|
|
24
|
+
|
|
25
|
+
Ask the user which mode they need:
|
|
26
|
+
|
|
27
|
+
| Mode | Trigger |
|
|
28
|
+
|---|---|
|
|
29
|
+
| **A: Create** | No existing file, or starting fresh |
|
|
30
|
+
| **B: Audit** | Has an existing agent.md / claude.md to optimize |
|
|
31
|
+
| **C: Band-Aid** | Quick fix for a specific agent mistake |
|
|
32
|
+
|
|
33
|
+
---
|
|
34
|
+
|
|
35
|
+
### 3. Execute
|
|
36
|
+
|
|
37
|
+
Follow the corresponding Mode (A, B, or C) from the skill instructions exactly.
|
|
38
|
+
|
|
39
|
+
---
|
|
40
|
+
|
|
41
|
+
### 4. Deliver
|
|
42
|
+
|
|
43
|
+
- Save the generated/optimized file to the user's preferred location
|
|
44
|
+
- Show before/after line count (Mode B)
|
|
45
|
+
- Confirm with user before overwriting any existing file
|
|
46
|
+
|
|
47
|
+
---
|
|
48
|
+
|
|
49
|
+
## When to Use
|
|
50
|
+
|
|
51
|
+
- **New project** — Create a minimal agent.md from scratch
|
|
52
|
+
- **Existing bloated file** — Audit and strip an overgrown context file
|
|
53
|
+
- **Agent keeps messing up** — Add a targeted Band-Aid rule
|
|
54
|
+
- **Cost optimization** — Reduce token overhead from verbose context files
|
|
@@ -17,7 +17,7 @@ Before writing ANY Remotion code, you MUST read the relevant rule files. Do NOT
|
|
|
17
17
|
### Step 0.1: Read the Main Skill
|
|
18
18
|
// turbo
|
|
19
19
|
```bash
|
|
20
|
-
cat "
|
|
20
|
+
cat ".agent/skills/remotion/SKILL.md"
|
|
21
21
|
```
|
|
22
22
|
|
|
23
23
|
### Step 0.2: Read REQUIRED Rule Files
|
|
@@ -27,62 +27,62 @@ Based on what the video needs, you MUST read the corresponding rule files. This
|
|
|
27
27
|
**For ANY animation work (ALWAYS READ THESE):**
|
|
28
28
|
// turbo
|
|
29
29
|
```bash
|
|
30
|
-
cat "
|
|
31
|
-
cat "
|
|
32
|
-
cat "
|
|
30
|
+
cat ".agent/skills/remotion/rules/animations.md"
|
|
31
|
+
cat ".agent/skills/remotion/rules/timing.md"
|
|
32
|
+
cat ".agent/skills/remotion/rules/sequencing.md"
|
|
33
33
|
```
|
|
34
34
|
|
|
35
35
|
**If using scene transitions:**
|
|
36
36
|
// turbo
|
|
37
37
|
```bash
|
|
38
|
-
cat "
|
|
38
|
+
cat ".agent/skills/remotion/rules/transitions.md"
|
|
39
39
|
```
|
|
40
40
|
|
|
41
41
|
**If using text animations (typewriter, highlights, etc):**
|
|
42
42
|
// turbo
|
|
43
43
|
```bash
|
|
44
|
-
cat "
|
|
45
|
-
cat "
|
|
44
|
+
cat ".agent/skills/remotion/rules/text-animations.md"
|
|
45
|
+
cat ".agent/skills/remotion/rules/fonts.md"
|
|
46
46
|
```
|
|
47
47
|
|
|
48
48
|
**If using images, videos, or audio:**
|
|
49
49
|
// turbo
|
|
50
50
|
```bash
|
|
51
|
-
cat "
|
|
52
|
-
cat "
|
|
53
|
-
cat "
|
|
54
|
-
cat "
|
|
51
|
+
cat ".agent/skills/remotion/rules/assets.md"
|
|
52
|
+
cat ".agent/skills/remotion/rules/images.md"
|
|
53
|
+
cat ".agent/skills/remotion/rules/videos.md"
|
|
54
|
+
cat ".agent/skills/remotion/rules/audio.md"
|
|
55
55
|
```
|
|
56
56
|
|
|
57
57
|
**If using 3D elements:**
|
|
58
58
|
// turbo
|
|
59
59
|
```bash
|
|
60
|
-
cat "
|
|
60
|
+
cat ".agent/skills/remotion/rules/3d.md"
|
|
61
61
|
```
|
|
62
62
|
|
|
63
63
|
**If using captions/subtitles:**
|
|
64
64
|
// turbo
|
|
65
65
|
```bash
|
|
66
|
-
cat "
|
|
67
|
-
cat "
|
|
66
|
+
cat ".agent/skills/remotion/rules/display-captions.md"
|
|
67
|
+
cat ".agent/skills/remotion/rules/transcribe-captions.md"
|
|
68
68
|
```
|
|
69
69
|
|
|
70
70
|
**If using charts/data visualization:**
|
|
71
71
|
// turbo
|
|
72
72
|
```bash
|
|
73
|
-
cat "
|
|
73
|
+
cat ".agent/skills/remotion/rules/charts.md"
|
|
74
74
|
```
|
|
75
75
|
|
|
76
76
|
**If using Lottie animations:**
|
|
77
77
|
// turbo
|
|
78
78
|
```bash
|
|
79
|
-
cat "
|
|
79
|
+
cat ".agent/skills/remotion/rules/lottie.md"
|
|
80
80
|
```
|
|
81
81
|
|
|
82
82
|
**If using GIFs:**
|
|
83
83
|
// turbo
|
|
84
84
|
```bash
|
|
85
|
-
cat "
|
|
85
|
+
cat ".agent/skills/remotion/rules/gifs.md"
|
|
86
86
|
```
|
|
87
87
|
|
|
88
88
|
> ⛔ **DO NOT PROCEED TO PHASE 1 UNTIL YOU HAVE READ THE RELEVANT RULES.**
|
|
@@ -5,7 +5,7 @@ description: Google Stitch design-to-code workflow. Routes to the correct sub-sk
|
|
|
5
5
|
# /stitch — Google Stitch Design Platform
|
|
6
6
|
|
|
7
7
|
## Overview
|
|
8
|
-
This workflow orchestrates the full Stitch design-to-code pipeline. All skills live in
|
|
8
|
+
This workflow orchestrates the full Stitch design-to-code pipeline. All skills live in `.agent/skills/stitch/`.
|
|
9
9
|
|
|
10
10
|
**The Stitch bundle is NOT a pick-one-and-go system.** It's a **pipeline** of 6 skills that chain together. Your job is to guide the user through the journey, always suggesting the next logical step.
|
|
11
11
|
|
|
@@ -13,7 +13,7 @@ This workflow orchestrates the full Stitch design-to-code pipeline. All skills l
|
|
|
13
13
|
// turbo
|
|
14
14
|
Read the master SKILL.md — it contains the **Pipeline Orchestration Protocol** which you MUST follow:
|
|
15
15
|
```
|
|
16
|
-
view_file
|
|
16
|
+
view_file .agent/skills/stitch/SKILL.md
|
|
17
17
|
```
|
|
18
18
|
|
|
19
19
|
## Step 2: Identify the Entry Point
|
|
@@ -35,7 +35,7 @@ Use the decision tree from the master SKILL.md to determine WHERE the user enter
|
|
|
35
35
|
// turbo
|
|
36
36
|
Read the specific sub-skill's `SKILL.md`:
|
|
37
37
|
```
|
|
38
|
-
view_file
|
|
38
|
+
view_file .agent/skills/stitch/{sub-skill}/SKILL.md
|
|
39
39
|
```
|
|
40
40
|
|
|
41
41
|
## Step 4: Check for Supporting Resources
|
|
@@ -46,7 +46,7 @@ Each sub-skill may have `resources/`, `examples/`, and `scripts/` folders:
|
|
|
46
46
|
- **scripts/**: Validation and networking helpers
|
|
47
47
|
|
|
48
48
|
```
|
|
49
|
-
ls
|
|
49
|
+
ls .agent/skills/stitch/{sub-skill}/
|
|
50
50
|
```
|
|
51
51
|
|
|
52
52
|
## Step 5: Execute the Sub-Skill
|
|
@@ -111,25 +111,14 @@ customModes:
|
|
|
111
111
|
|
|
112
112
|
**MANDATORY STEP — DO NOT SKIP.**
|
|
113
113
|
|
|
114
|
-
|
|
115
|
-
are available for sub-agents:
|
|
114
|
+
**Your system prompt lists all available skills with their full absolute paths.** Check there FIRST — do NOT guess paths.
|
|
116
115
|
|
|
116
|
+
If your system prompt doesn't list them, fall back to scanning the filesystem:
|
|
117
117
|
```powershell
|
|
118
|
-
# Scan for skills — check all known locations
|
|
119
|
-
# Project-local skills
|
|
120
118
|
ls .agent/skills/ 2>$null
|
|
121
|
-
|
|
122
|
-
# Global skills (path varies by AI client)
|
|
123
|
-
ls ~/.gemini/antigravity/skills/ 2>$null # Gemini / Antigravity
|
|
124
|
-
ls ~/.kilocode/skills/ 2>$null # KiloCode
|
|
125
|
-
# Add other client paths as discovered
|
|
126
|
-
|
|
127
|
-
# For each skill found, read the SKILL.md to understand its purpose
|
|
128
|
-
# Focus on the name, description, and triggers
|
|
129
119
|
```
|
|
130
120
|
|
|
131
|
-
|
|
132
|
-
The agent's system prompt may also list available skills — use that too.
|
|
121
|
+
For each relevant skill found, read its SKILL.md to understand its purpose.
|
|
133
122
|
|
|
134
123
|
Create a skills registry for this session:
|
|
135
124
|
|
|
@@ -313,33 +302,25 @@ customModes:
|
|
|
313
302
|
### Workflow to Follow
|
|
314
303
|
> **Primary Workflow:** `/vibe-build` (or whichever is assigned)
|
|
315
304
|
>
|
|
316
|
-
>
|
|
317
|
-
>
|
|
305
|
+
> Read the workflow from your system prompt — it lists all workflows with
|
|
306
|
+
> their full paths. Use `view_file` on the path listed there.
|
|
307
|
+
> Fallback: `cat .agent/workflows/vibe-build.md`
|
|
308
|
+
> Fallback: `cat .agent/global_workflows/vibe-build.md`
|
|
318
309
|
|
|
319
310
|
### Load Prime Agent Context
|
|
320
311
|
> **MANDATORY:** Run `/vibe-primeAgent` before starting ANY work.
|
|
321
312
|
> This loads coding guidelines, project state, and current work.
|
|
322
313
|
|
|
323
314
|
### Required Skills
|
|
324
|
-
> **
|
|
315
|
+
> **Your system prompt lists all skills with their absolute paths.**
|
|
316
|
+
> Look up each skill below in your system prompt and `view_file` its SKILL.md.
|
|
325
317
|
>
|
|
326
|
-
> | Skill |
|
|
327
|
-
>
|
|
328
|
-
> | `nextjs-standards` |
|
|
329
|
-
> | `ai-sdk` |
|
|
318
|
+
> | Skill | Why |
|
|
319
|
+
> |-------|-----|
|
|
320
|
+
> | `nextjs-standards` | Project uses Next.js App Router |
|
|
321
|
+
> | `ai-sdk` | This task involves AI features |
|
|
330
322
|
>
|
|
331
|
-
>
|
|
332
|
-
> - `.agent/skills/` (project-local)
|
|
333
|
-
> - `~/.gemini/antigravity/skills/` (Antigravity)
|
|
334
|
-
> - `~/.kilocode/skills/` (KiloCode)
|
|
335
|
-
> - Or check your system prompt for available skills
|
|
336
|
-
>
|
|
337
|
-
> **How to load:** `cat [path-to-skill]/SKILL.md` and follow the skill's instructions.
|
|
338
|
-
|
|
339
|
-
### Check Available Skills
|
|
340
|
-
> Before starting, scan all known skills directories for any additional
|
|
341
|
-
> skills that may help with this specific task. If you find a relevant skill
|
|
342
|
-
> not listed above, load it.
|
|
323
|
+
> Fallback (if not in system prompt): `ls .agent/skills/`
|
|
343
324
|
|
|
344
325
|
---
|
|
345
326
|
|
|
@@ -148,25 +148,15 @@ customModes:
|
|
|
148
148
|
|
|
149
149
|
**CRITICAL: Before making any decisions, scan available skills and workflows.**
|
|
150
150
|
|
|
151
|
-
|
|
152
|
-
Check ALL known locations and use whichever exists:
|
|
151
|
+
**Your system prompt lists all available skills and workflows with their full absolute paths.** Check there FIRST — do NOT guess paths.
|
|
153
152
|
|
|
153
|
+
Fallback (if not in system prompt):
|
|
154
154
|
```powershell
|
|
155
|
-
# Scan for skills — check all known locations
|
|
156
|
-
# Project-local skills
|
|
157
155
|
ls .agent/skills/ 2>$null
|
|
158
|
-
|
|
159
|
-
# Global skills (path varies by AI client)
|
|
160
|
-
ls ~/.gemini/antigravity/skills/ 2>$null # Gemini / Antigravity
|
|
161
|
-
ls ~/.kilocode/skills/ 2>$null # KiloCode
|
|
162
|
-
# Add other client paths as discovered
|
|
163
|
-
|
|
164
|
-
# Scan available workflows (always project-local)
|
|
165
156
|
ls .agent/workflows/ 2>$null
|
|
157
|
+
ls .agent/global_workflows/ 2>$null
|
|
166
158
|
```
|
|
167
159
|
|
|
168
|
-
**Use whichever skills directory exists.** If multiple exist, merge the lists.
|
|
169
|
-
The agent's system prompt may also list available skills — use that too.
|
|
170
160
|
These will be passed to the Orchestrator for injection into sub-agent tasks.
|
|
171
161
|
|
|
172
162
|
---
|