skill-forge-convert

Name: skill-forge-convert
Author: AgriciDaniel

78stars33forks

How to install skill-forge-convert?

agricidaniel/skill-forge/skill-forge-convert

$npx -y skills add AgriciDaniel/skill-forge --skill skill-forge-convert

Installs into the current project.

Summary

Convert Claude Code skills to work on OpenAI Codex, Google Gemini CLI, Google Antigravity, and Cursor. Analyzes platform-specific features, generates target files (openai.yaml, AGENTS.md, GEMINI.md, .mdc rules), adapts frontmatter, converts MCP config, and produces compatibility

Files · 1

View on GitHub

SKILL.md

1	# Skill Conversion — Multi-Platform
2
3	Convert Claude Code skills to work on OpenAI Codex, Google Gemini CLI, Google
4	Antigravity, and Cursor while maintaining quality and following each platform's
5	best practices.
6
7	## Process
8
9	### Step 1: Read Source Skill
10
11	1. Read the source `SKILL.md` and parse frontmatter
12	2. Detect skill complexity tier (1-4)
13	3. Inventory all files: scripts/, references/, agents/, sub-skills
14	4. Identify MCP config (`.mcp.json`) if present
15
16	### Step 2: Analyze Platform Compatibility
17
18	Run dry-run analysis first:
19	```bash
20	python scripts/convert_skill.py <path> --dry-run --target all
21	```
22
23	Review the compatibility report:
24	- Portable fields: Transfer directly (name, description, license)
25	- Adaptable fields: Need platform-specific handling
26	- Claude-only fields: Will be stripped with warnings
27
28	### Step 3: Convert to Target Platforms
29
30	Run the conversion:
31	```bash
32	python scripts/convert_skill.py <path> --target codex,gemini,antigravity,cursor --output dist/
33	```
34
35	For MCP config conversion:
36	```bash
37	python scripts/convert_skill.py <path> --target all --output dist/ --include-mcp
38	```
39
40	### Step 4: Handle Claude-Only Features
41
42	Features that need manual adaptation per platform:
43
44	\| Claude Feature \| Codex \| Gemini/Antigravity \| Cursor \|
45	\|---------------\|-------\|-------------------\|--------\|
46	\| `allowed-tools` \| Supported \| Remove; all tools available \| Remove; no equivalent \|
47	\| `context: fork` \| No equivalent \| No equivalent \| No equivalent \|
48	\| `hooks` \| 1 event (notify) \| Partial (CLI) \| 6 events (beta) \|
49	\| `model` selection \| No equivalent \| No equivalent \| No equivalent \|
50	\| `Task` delegation \| Break into separate skills \| Break into separate skills \| Background Agents (Ultra plan) \|
51	\| `/slash` commands \| `$mention` syntax \| Description-based activation \| Description-based + `@rule` \|
52	\| Sub-skill routing \| Separate skills with `$mention` \| Separate skills, LLM-routed \| Separate skills, LLM-routed \|
53
54	### Step 5: Validate Converted Skills
55
56	Run validation on each generated output:
57	```bash
58	python scripts/validate_skill.py dist/codex/<skill-name>/
59	python scripts/validate_skill.py dist/gemini/<skill-name>/
60	python scripts/validate_skill.py dist/antigravity/<skill-name>/
61	python scripts/validate_skill.py dist/cursor/<skill-name>/
62	```
63
64	Fix any critical or high-priority issues before proceeding.
65
66	### Step 6: Generate Deployment Report
67
68	Present the user with a summary:
69
70	```
71	## Conversion Report: {skill-name}
72
73	\| Platform \| Score \| Files \| Warnings \| Manual Steps \|
74	\|----------\|-------\|-------\|----------\|-------------\|
75	\| Codex \| 92% \| 4 \| 2 \| 1 \|
76	\| Gemini \| 88% \| 3 \| 3 \| 1 \|
77	\| Antigravity \| 88% \| 3 \| 3 \| 1 \|
78	\| Cursor \| 88% \| 3 \| 3 \| 1 \|
79
80	### Generated Files
81	[list per platform]
82
83	### Warnings
84	[list per platform]
85
86	### Manual Steps Required
87	[list per platform]
88	```
89
90	### Step 7: Generate Multi-Platform Install Script
91
92	When converting for all platforms (`--target all`), the script auto-generates
93	`install-multiplatform.sh` that:
94	- Auto-detects the current agent platform
95	- Installs to the correct skill path
96	- Supports `--platform` flag for explicit selection
97
98	## Platform Quick Reference
99
100	\| Platform \| Skill Path \| Instruction File \| Config Format \|
101	\|----------\|-----------\|-----------------\|---------------\|
102	\| Claude Code \| `.claude/skills/` \| `CLAUDE.md` \| JSON (`.mcp.json`) \|
103	\| OpenAI Codex \| `.agents/skills/` \| `AGENTS.md` \| TOML (`config.toml`) \|
104	\| Gemini CLI \| `.gemini/skills/` \| `GEMINI.md` \| JSON (`settings.json`) \|
105	\| Antigravity \| `.agent/skills/` \| `GEMINI.md` \| JSON (`mcp_config.json`) \|
106	\| Cursor \| `.cursor/skills/` \| `.cursor/rules/*.mdc` \| JSON (`mcp.json`) \|
107
108	For full platform specs, load `references/platforms.md`.
109
110	## Tier Conversion Guidelines
111
112	- Tier 1 (single SKILL.md): Converts cleanly. Auto-convert recommended.
113	- Tier 2 (skill + scripts): Converts well. Verify script paths.
114	- Tier 3 (multi-skill): Partial. Each sub-skill converts independently. Routing needs manual work.
115	- Tier 4 (ecosystem): Manual review required. Subagent delegation needs platform adaptation.

People who liked this also viewed

6 picks

Type

skill-forge-convert

78stars33forks

View on GitHub

How to install skill-forge-convert?

agricidaniel/skill-forge/skill-forge-convert

$npx -y skills add AgriciDaniel/skill-forge --skill skill-forge-convert

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Skill Conversion — Multi-Platform
2
3	Convert Claude Code skills to work on OpenAI Codex, Google Gemini CLI, Google
4	Antigravity, and Cursor while maintaining quality and following each platform's
5	best practices.
6
7	## Process
8
9	### Step 1: Read Source Skill
10
11	1. Read the source `SKILL.md` and parse frontmatter
12	2. Detect skill complexity tier (1-4)
13	3. Inventory all files: scripts/, references/, agents/, sub-skills
14	4. Identify MCP config (`.mcp.json`) if present
15
16	### Step 2: Analyze Platform Compatibility
17
18	Run dry-run analysis first:
19	```bash
20	python scripts/convert_skill.py <path> --dry-run --target all
21	```
22
23	Review the compatibility report:
24	- Portable fields: Transfer directly (name, description, license)
25	- Adaptable fields: Need platform-specific handling
26	- Claude-only fields: Will be stripped with warnings
27
28	### Step 3: Convert to Target Platforms
29
30	Run the conversion:
31	```bash
32	python scripts/convert_skill.py <path> --target codex,gemini,antigravity,cursor --output dist/
33	```
34
35	For MCP config conversion:
36	```bash
37	python scripts/convert_skill.py <path> --target all --output dist/ --include-mcp
38	```
39
40	### Step 4: Handle Claude-Only Features
41
42	Features that need manual adaptation per platform:
43
44	\| Claude Feature \| Codex \| Gemini/Antigravity \| Cursor \|
45	\|---------------\|-------\|-------------------\|--------\|
46	\| `allowed-tools` \| Supported \| Remove; all tools available \| Remove; no equivalent \|
47	\| `context: fork` \| No equivalent \| No equivalent \| No equivalent \|
48	\| `hooks` \| 1 event (notify) \| Partial (CLI) \| 6 events (beta) \|
49	\| `model` selection \| No equivalent \| No equivalent \| No equivalent \|
50	\| `Task` delegation \| Break into separate skills \| Break into separate skills \| Background Agents (Ultra plan) \|
51	\| `/slash` commands \| `$mention` syntax \| Description-based activation \| Description-based + `@rule` \|
52	\| Sub-skill routing \| Separate skills with `$mention` \| Separate skills, LLM-routed \| Separate skills, LLM-routed \|
53
54	### Step 5: Validate Converted Skills
55
56	Run validation on each generated output:
57	```bash
58	python scripts/validate_skill.py dist/codex/<skill-name>/
59	python scripts/validate_skill.py dist/gemini/<skill-name>/
60	python scripts/validate_skill.py dist/antigravity/<skill-name>/
61	python scripts/validate_skill.py dist/cursor/<skill-name>/
62	```
63
64	Fix any critical or high-priority issues before proceeding.
65
66	### Step 6: Generate Deployment Report
67
68	Present the user with a summary:
69
70	```
71	## Conversion Report: {skill-name}
72
73	\| Platform \| Score \| Files \| Warnings \| Manual Steps \|
74	\|----------\|-------\|-------\|----------\|-------------\|
75	\| Codex \| 92% \| 4 \| 2 \| 1 \|
76	\| Gemini \| 88% \| 3 \| 3 \| 1 \|
77	\| Antigravity \| 88% \| 3 \| 3 \| 1 \|
78	\| Cursor \| 88% \| 3 \| 3 \| 1 \|
79
80	### Generated Files
81	[list per platform]
82
83	### Warnings
84	[list per platform]
85
86	### Manual Steps Required
87	[list per platform]
88	```
89
90	### Step 7: Generate Multi-Platform Install Script
91
92	When converting for all platforms (`--target all`), the script auto-generates
93	`install-multiplatform.sh` that:
94	- Auto-detects the current agent platform
95	- Installs to the correct skill path
96	- Supports `--platform` flag for explicit selection
97
98	## Platform Quick Reference
99
100	\| Platform \| Skill Path \| Instruction File \| Config Format \|
101	\|----------\|-----------\|-----------------\|---------------\|
102	\| Claude Code \| `.claude/skills/` \| `CLAUDE.md` \| JSON (`.mcp.json`) \|
103	\| OpenAI Codex \| `.agents/skills/` \| `AGENTS.md` \| TOML (`config.toml`) \|
104	\| Gemini CLI \| `.gemini/skills/` \| `GEMINI.md` \| JSON (`settings.json`) \|
105	\| Antigravity \| `.agent/skills/` \| `GEMINI.md` \| JSON (`mcp_config.json`) \|
106	\| Cursor \| `.cursor/skills/` \| `.cursor/rules/*.mdc` \| JSON (`mcp.json`) \|
107
108	For full platform specs, load `references/platforms.md`.
109
110	## Tier Conversion Guidelines
111
112	- Tier 1 (single SKILL.md): Converts cleanly. Auto-convert recommended.
113	- Tier 2 (skill + scripts): Converts well. Verify script paths.
114	- Tier 3 (multi-skill): Partial. Each sub-skill converts independently. Routing needs manual work.
115	- Tier 4 (ecosystem): Manual review required. Subagent delegation needs platform adaptation.

People who liked this also viewed

6 picks

Type