package
Skill25 starsChanged 7 months agoMITView on GitHub
Package a product for distribution across 33+ platforms. Strips WHY comments, generates triple manifests (vault.yaml + plugin.json + agentskills.yaml), stages .publish/. Use when: 'package', 'prepare for publish', or 'bundle it'.
- Reads credentials
4 other projects use a version of this.
Adopted by
Most recently taken up by
- zhu-xiaowei/Batontheir own version, first seen Oct 13, 2026
- apache/skywalkingtheir own version, first seen Oct 8, 2026
- jerry-ai-dev/MODULAR-RAG-MCP-SERVERtheir own version, first seen Oct 8, 2026
- stardots-io/stardots-backup-skilltheir own version, first seen Mar 11, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- apache/skywalking★ 25ksame name
- jerry-ai-dev/MODULAR-RAG-MCP-SERVER★ 1.2ksame name
- zhu-xiaowei/Baton★ 14same name
- stardots-io/stardots-backup-skill★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
Show its rules and text
We split this skill into the rules it gives an agent
35 rules. About 0 are common, 12 are distinctive, and 23 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈12
Distinctive · estimate
This author's own choices. Worth a look.
23
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
When to use: After /validate passes. Before /publish.
- Distinctive: line 22
When NOT to use: If the product hasn't been validated yet (run /validate first).
- Very specific: line 33
Read .meta.yaml → verify state is "validated" and MCS score >= 75%
- Very specific: line 34
Read creator.yaml → load author metadata for manifests.
- Very specific: line 56
Read .meta.yaml for validation state:
- Distinctive: line 57
If state != "validated" AND state != "packaged": "Product not validated.
- Very specific: line 58
If MCS score < 75%: "Product below MCS-1 threshold.
- Distinctive: line 59
If mcs_target is "MCS-2" or "MCS-3" AND (test_result is null OR test_result != "pass"): "MCS-2+ products require behavioral testing before …
- Very specific: line 63
Check product-dna/{type}.yaml for package_rename field.
- Distinctive: line 73
Create a clean copy of all product files.
- Very specific: line 87
Add a marketplace_url field to the YAML frontmatter of the primary file.
- Very specific: line 97
Add marketplace_url as a NEW field — never modify the existing description
- Very specific: line 98
Only add to the .publish/ copy — never touch the original in workspace/
- Very specific: line 99
Use the slug from vault.yaml
- Very specific: line 111
Never modify originals — work on the copy in .publish/.
- Very specific: line 125
Precondition: Verify .publish/README.md exists before injection.
- Very specific: line 133
Add a blank line before the --- separator to avoid markdown parsing issues
- Distinctive: line 141
After WHY removal, verify the stripped files are still structurally valid:
- Distinctive: line 145
If any check fails: report the specific file + line + issue.
- Very specific: line 149
Read .meta.yaml → intent_declaration.language (written by /create Step 11).
- Very specific: line 181
Read .meta.yaml → intelligence fields.
- Distinctive: line 194
If the product has a non-zero price:
- Very specific: line 196
Stripe check: Run myclaude stripe status 2>/dev/null
- Very specific: line 197
If not connected: BLOCKING — "Paid products require Stripe.
- Very specific: line 200
Competitive pricing scan: Run myclaude search --category {type} --sort price-desc --limit 5 --json 2>/dev/null If results found, display:
- Distinctive: line 201
If results found, display:
- Very specific: line 210
MCS-price alignment: If price > $5 AND mcs_level < 2, warn: "Price is ${price} but MCS is level {level}.
- Very specific: line 249
Only agent key is valid inside settings.json plugin settings (not skill, workflow, etc.)
- Very specific: line 314
Only generate for types that map to Agent Skills format (skill, agent, prompt, workflow).
- Very specific: line 336
If any excluded-for-secrets file is found, WARN: "Sensitive file {file} excluded from package.
- Very specific: line 340
Run the same secrets scan patterns (from config.yaml → secret_content_patterns) against BOTH .publish/ AND workspace/{slug}/ originals.
- Distinctive: line 372
If any FAIL: abort with message listing missing artifacts and expected structure.
- Very specific: line 376
Run Stage 1 (structural) + Stage 6 (CLI preflight) on .publish/ contents.
- Distinctive: line 377
If any check fails, report and abort — don't leave broken package.
- Distinctive: line 449
When context is compressed, preserve:
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Packager
Stage a product for distribution with triple manifests (MyClaude + Anthropic Plugin + Agent Skills universal).
When to use: After /validate passes. Before /publish.
When NOT to use: If the product hasn't been validated yet (run /validate first).
Activation Protocol
- Shared preamble: Load
references/quality/activation-preamble.md— context assembly, persona adaptation, deterministic routing rules. - Identify product:
$ARGUMENTSas slug →workspace/{slug}/1b. Mode selection (Express vs Guided). Readcreator.yaml → preferences.workflow_style. Resolve the flow mode:--expressflag ORworkflow_style == "autonomous"→ Express mode. Skip the "confirm before stage" prompt, skip the manifest preview step, and produce a single post-stage summary. The safety invariants below still hold — only the conversational confirmations are trimmed.workflow_style == "guided"or missing → Guided mode (default). Confirm before staging, show manifest preview, and walk the creator through the pipeline with full voice.
- Read
.meta.yaml→ verify state is "validated" and MCS score >= 75% - Read
creator.yaml→ load author metadata for manifests. If missing → "Creator profile not found. Run/onboardfirst." and stop. - Maintain creator persona: Adapt language, depth, and examples to
profile.typeandtechnical_levelthroughout this skill's execution. A developer gets code examples; a domain expert gets plain language. - Load
product-dna/{type}.yaml→ get install_target - Load
config.yaml→ vault_defaults for missing fields 6b. Load proactives: Loadreferences/engine-proactive.md— wire #1 (pipeline guidance: after package, guide to /publish), #20 (test mandate: if MCS-2+ and.meta.yaml.test_result != "pass", block packaging with message "Run /test first"). - Load voice identity: Load
references/quality/engine-voice-core.md. Every user-facing line in this skill honors the ✦ signature, three tones, and six anti-patterns. - CLI contract: Load
references/cli-contract.mdfor unified error handling. This skill has conditional severity based on pricing model. Severity map:- Blocking (paid products only):
stripe status— if product price > 0 and Stripe not connected, halt packaging - Silent-skip (free products):
stripe status— irrelevant for free products, skip without warning - Silent-skip:
search --category {type} --sort price-desc— competitive pricing scan, skip without warning on failure - All queries: append
2>/dev/null, 15s timeout, treat invalid JSON as CLI absent
- Blocking (paid products only):
Core Instructions
PACKAGING PIPELINE
SAFETY INVARIANTS — see end of file (positioned per D19 attention-aware authoring for maximum model compliance).
Step 1 — Verify Validation
Read .meta.yaml for validation state:
- If state != "validated" AND state != "packaged": "Product not validated. Run /validate first."
- If MCS score < 75%: "Product below MCS-1 threshold. Run /validate --fix."
- If mcs_target is "MCS-2" or "MCS-3" AND (test_result is null OR test_result != "pass"): "MCS-2+ products require behavioral testing before packaging. Run /test first." — BLOCKING. This gate ensures structural validation (/validate) AND behavioral validation (/test) both pass before distribution.
Step 1.5 — Package Rename (if applicable)
Check product-dna/{type}.yaml for package_rename field. If present, rename the
primary file to the specified pattern during copy to .publish/.
Currently applies to:
- minds:
AGENT.md→{slug}.md(flat file convention for.claude/agents/)
This rename ensures the installed filename matches Claude Code's native discovery pattern.
Step 2 — Strip WHY Comments + Inject Attribution (SE-D17)
Create a clean copy of all product files.
REMOVE:
<!-- WHY: ... -->(HTML comment blocks — may span multiple lines)- Lines matching
# WHY:pattern (markdown comment format) # WHY:pattern in shell scripts (.shfiles)- JSON keys prefixed with
_(e.g.,_comment,_why,_events_reference) from.jsonfiles. These are documentation keys used in workspace but must not pollute the user's settings.json when hooks are installed. Strip by parsing JSON, removing_-prefixed top-level keys, and re-serializing.
INJECT TWO THINGS into the .publish/ copy of the primary file (SKILL.md, AGENT.md, etc.):
1. Marketplace URL in frontmatter (PRD D3 — viral vector #1):
Add a marketplace_url field to the YAML frontmatter of the primary file. This field is read by Claude Code in EVERY session where the product is installed — making every buyer a passive distribution carrier.
---
name: {slug}
description: >
{original description — DO NOT MODIFY}
marketplace_url: "https://myclaude.sh/p/{slug}"
---Rules:
- Add
marketplace_urlas a NEW field — never modify the existingdescription - Only add to the
.publish/copy — never touch the original in workspace/ - Use the slug from vault.yaml
2. Attribution comment at end of file:
<!-- Published on MyClaude (myclaude.sh) | Quality: MCS-{level} ({score}%) | Engine: Studio v3 -->This attribution comment is:
- Invisible to Claude (HTML comment, does not affect behavior)
- Visible in source code (GitHub renders it in raw view)
- Indexable by search engines (creates discoverable MyClaude references)
- A quality signal (MCS score travels with the product)
Never modify originals — work on the copy in .publish/.
Step 2c — Inject README Badges (PRD B2 — viral vectors #2)
Append MCS and Available badges to the end of .publish/README.md:
---
[](https://myclaude.sh/quality)
[](https://myclaude.sh/p/{slug})
<sub>Built with MyClaude Studio Engine</sub>Precondition: Verify .publish/README.md exists before injection. If missing (bundle type has no README requirement), skip badge injection silently — do not create a README just for badges.
README quality gate: Before badge injection, verify README follows templates/readme/README.md.template structure — at minimum: Hero, Install (with myclaude install {slug}), Quick Start, Features (3+), Requirements. If missing critical sections, WARN: "README is missing sections that help users decide to install. Consider running /fill to complete it." Continue packaging — this is advisory, not blocking.
Rules:
- Inject ONLY in
.publish/README.md— never modify the original - MCS level from
.meta.yamlstate.dna_complianceoroverall_score. Default to 1 if not validated - Slug from vault.yaml
namefield - Add a blank line before the
---separator to avoid markdown parsing issues
Step 2b — Verify WHY Removal
Grep all files in .publish/ for the pattern WHY:. If match count > 0, re-run stripping on the matched files. If still > 0 after second pass, report error with file paths and line numbers — do not proceed to manifest generation with unstripped WHY comments.
Step 2d — Post-Strip Integrity Check
After WHY removal, verify the stripped files are still structurally valid:
- For
.mdfiles: verify YAML frontmatter still parses (regex for^---\n...\n---\n) - For
.jsonfiles (hooks.json, settings-fragment.json): parse with JSON validator after_-prefixed key removal. If parse fails → abort with file path and error. Never ship invalid JSON. - For
.shfiles: verify shebang line survived stripping (first line must start with#!) - If any check fails: report the specific file + line + issue. Do NOT proceed to manifest generation.
Step 3 — Generate vault.yaml
Read .meta.yaml → intent_declaration.language (written by /create Step 11). This is
the canonical source of truth for the product's source language — it captures what
language the product was forged in, not what language the creator is currently
using. If the field is missing (pre-Wave-3 legacy product), fall back to
creator.yaml → creator.language, then to "en" as the final default.
# REQUIRED (from .meta.yaml + creator.yaml)
name: "{slug}"
version: "{from .meta.yaml or config default}"
type: "{product type}"
description: "{from README.md first paragraph or .meta.yaml}"
entry: "{primary file from product-dna/{type}.yaml}"
license: "{from creator.yaml or config default}"
price: {from .meta.yaml or 0}
tags: ["{from .meta.yaml}"]
# LOCALE CONTRACT — see references/locale-adaptive-clause.md §8 for the full wiring map.
# source_language travels with the product to the marketplace so the UI can
# render the "Source: {language} · Adapts to your language at runtime" badge.
# These four fields are REQUIRED for any product that was forged with the
# locale-adaptive clause (all 6 certified types). Omit the block only for
# types that do not carry the clause (bundle, claude-md, application, hooks,
# statusline, design-system, workflow — per locale-adaptive-clause.md §4).
source_language: "{from .meta.yaml intent_declaration.language, fallback to creator.yaml language, fallback to 'en'}"
locale_adaptive: true
locale_adaptive_source: "references/locale-adaptive-clause.md"
locale_adaptive_version: "1.0"Step 3b — Pricing Intelligence (Intelligence Layer integration)
Read .meta.yaml → intelligence fields. If value_score is populated (set by /validate Stage 8), display value-informed pricing context:
Value Intelligence:
Value score: {value_score}/12 ({pricing_strategy})
Suggested range: ${suggested_price_range[0]}-${suggested_price_range[1]}
Current price: ${price from .meta.yaml or vault_defaults}
{If price == 0 AND value_score >= 5: "Your product scores {value_score}/12 — consider pricing at ${suggested_price_range[0]}+ based on depth and uniqueness."}
{If price > 0 AND price > suggested_price_range[1]: "Current price exceeds value signal range. Ensure README clearly communicates the premium."}
{If price > 0 AND price < suggested_price_range[0]: "Priced below value signal. You're leaving value on the table — or strategically building audience."}
{If value_score not set: "Run /validate --level=2 to compute value intelligence."}If the product has a non-zero price:
- Stripe check: Run
myclaude stripe status 2>/dev/null- If not connected: BLOCKING — "Paid products require Stripe. Run
myclaude stripe connectfirst." - If connected: proceed
- If not connected: BLOCKING — "Paid products require Stripe. Run
- Competitive pricing scan: Run
myclaude search --category {type} --sort price-desc --limit 5 --json 2>/dev/nullIf results found, display:Pricing context for {type} products: Highest: ${max_price} ({name}) Average: ${avg_price} Your price: ${price} {If price > avg: "Premium positioning" | If price < avg: "Value positioning" | If price == 0 and others charge: "Consider charging — similar products average ${avg}"} - MCS-price alignment: If price > $5 AND mcs_level < 2, warn:
"Price is ${price} but MCS is level {level}. Users expect MCS-2+ for paid products. Consider running
/validate --level=2to increase quality."
Epistemic caveat (always shown when value intelligence is displayed):
"Value signal is estimated from structural quality + market position. Real value is confirmed by daily use."
Continue to enrichment:
# ENRICHMENT (from .meta.yaml + creator.yaml + computed)
displayName: "{humanized from name}"
mcsLevel: {from last validation score}
language: "{from creator.yaml}"
longDescription: "{from README.md}"
readme: "README.md"
installTarget: "{from product-dna/{type}.yaml}"
language: "{from creator.yaml or 'en'}"
compatibility:
claude_code: ">=1.0.0"
# DISTRIBUTION DNA (viral growth layer — links back to MyClaude ecosystem)
engine: "myclaude-studio-engine"
marketplace: "https://myclaude.sh/p/{slug}"
badges:
mcs: "https://myclaude.sh/badge/mcs/{mcs_level_number}.svg"
available: "https://myclaude.sh/badge/available.svg"Step 4 — Generate plugin.json (Anthropic Plugin Marketplace)
Settings priority: Plugin settings are LOWEST priority in CC's 6-level hierarchy (Plugin < User < Project < Local < Flag < Policy). This means user/project settings always override plugin defaults. Hooks from plugins ACCUMULATE (never replace existing hooks). MCP servers from plugins REPLACE entirely.
Schema constraints (enforce strictly — wrong structure causes silent install failures):
authorMUST be an object{ "name": "...", "url": "..." }— never a plain stringskillsfield uses a directory path (e.g.,".claude/skills/slug/") — CC loads all .md files in itagentsfield uses an explicit file path array (e.g.,[".claude/skills/slug/AGENT.md"]) — NOT a directorycommandsfield uses a directory path (e.g.,".claude/commands/")hookskey does NOT belong in plugin.json — hooks are declared in a separatehooks.jsonfile- Only
agentkey is valid insidesettings.jsonplugin settings (notskill,workflow, etc.)
Emit only the fields relevant to the product type:
{
"name": "{slug}",
"description": "{description}",
"version": "{version}",
"author": { "name": "{from creator.yaml display_name}", "url": "https://myclaude.sh/u/{from creator.yaml username}" },
"license": "{license}",
"homepage": "https://myclaude.sh/p/{slug}",
// --- Include ONE of the following blocks based on product type ---
// type=skill or type=workflow or type=squad or type=claude-md:
"skills": ".claude/skills/{slug}/",
// type=agent:
"agents": [".claude/skills/{slug}/AGENT.md"],
// type=minds:
"agents": [".claude/agents/{slug}.md"],
// type=system (if it exposes commands):
"commands": ".claude/commands/"
// NOTE: hooks type — no content field in plugin.json. hooks.json is the artifact.
// NOTE: statusline type — no content field in plugin.json. statusline.sh is the artifact.
// NOTE: design-system, application, bundle — no content field; vault.yaml handles install.
}Step 4b — Generate agentskills.yaml (Agent Skills Universal — 33+ platforms)
The Agent Skills format (YAML frontmatter + markdown instructions) is the structured skill definition format used natively by Claude Code and recognized by other AI coding tools. Generate this manifest for cross-platform compatibility.
# Agent Skills Universal Manifest
name: "{slug}"
version: "{version}"
description: "{description}"
type: "{product type}"
author: "{from creator.yaml}"
license: "{license}"
entry: "{primary file}"
platforms:
- claude-code
- cursor
- codex
- gemini-cli
- copilot
- "*" # Compatible with any Agent Skills-compliant platform
install:
target: "{from product-dna install_target}"
files:
- "{list all files in .publish/ relative paths}"
metadata:
mcs_level: "{from validation}"
mcs_score: "{from validation}"
marketplace: "https://myclaude.sh/p/{slug}"Only generate for types that map to Agent Skills format (skill, agent, prompt, workflow). For types that don't map cleanly (design-system, application), generate only vault.yaml + plugin.json.
Step 5 — Calculate Checksum
SHA-256 of the entire .publish/ directory contents.
Step 6 — Stage .publish/
Create workspace/{slug}/.publish/ with:
- Cleaned product files (WHY comments stripped)
- vault.yaml (MyClaude marketplace)
- plugin.json (Anthropic plugin marketplace)
- agentskills.yaml (Agent Skills universal — if applicable type)
- CHANGELOG.md (generate if missing)
- LICENSE.md file (generate from license field — use .md extension, CLI rejects extensionless files)
EXCLUDE from .publish/ (internal Engine files — never distribute):
.meta.yaml(Engine product state tracking)domain-map.md(creator's working notes)- Any file starting with
.(hidden files, including.env) - Files matching secret patterns:
*.pem,*.key,*.p12,credentials*.json,*.env - If any excluded-for-secrets file is found, WARN: "Sensitive file
{file}excluded from package. If this is intentional, rename it."
Step 6c — Double-Source Secrets Scan
Run the same secrets scan patterns (from config.yaml → secret_content_patterns) against BOTH .publish/ AND workspace/{slug}/ originals. Compare results. If .publish/ has MORE matches than originals (stripping logic injected content), this is a CRITICAL error — abort packaging immediately and report the discrepancy.
Step 6b — Verify Type-Specific Install Artifacts
The CLI install command has type-specific handlers. The tarball MUST contain the artifacts
each handler expects, or install will silently skip critical operations.
| Type | Required in .publish/ | CLI Install Behavior |
| hooks | hooks.json + scripts/ (with .sh files) | Merges hooks.json into settings.local.json + copies scripts/ |
| statusline | statusline.sh + settings-fragment.json | Copies script + merges statusLine config into settings.local.json |
| minds | {slug}.md (renamed from AGENT.md in Step 1.5) | Copies flat file to .claude/agents/{slug}.md |
Verification (MUST pass or abort):
For hooks:
hooks.jsonexists in.publish/→ FAIL if missinghooks.jsonis valid JSON with top-level"hooks"key → FAIL if malformedscripts/directory exists with at least one.shfile → FAIL if missing- JSON
_-prefixed keys were stripped in Step 2 (e.g.,_comment,_events_reference)
For statusline:
statusline.shexists in.publish/→ FAIL if missingstatusline.shhas shebang#!/usr/bin/env bash→ WARN if missingsettings-fragment.jsonexists with"statusLine"key → FAIL if missing
For minds:
{slug}.mdexists in.publish/(Step 1.5 should have renamed it) → FAIL if missing- File has YAML frontmatter with
denied-tools→ WARN if missing (advisory-only guarantee)
For all other types: no additional verification (default install path handles them).
If any FAIL: abort with message listing missing artifacts and expected structure.
Step 7 — Re-validate
Run Stage 1 (structural) + Stage 6 (CLI preflight) on .publish/ contents.
If any check fails, report and abort — don't leave broken package.
Step 8 — Update State
# .meta.yaml updates
state:
phase: "packaged"
packaged_at: "{ISO timestamp}"Output Format
UX Stack (load before rendering output):
references/ux-experience-system.md§1 Context Assembly + §2.2 Archetype-Aware Insights (pricing emphasis per creator goals)references/ux-vocabulary.md— translate all termsreferences/quality/engine-voice.md— Brand DNA
Cognitive rendering: /package is a transition moment — the product is ready to meet the world. Adapt output: creators with revenue goals get pricing intelligence front-and-center. Community-focused creators get distribution channel emphasis. Expert creators get compact technical summary. Beginners get reassurance: "Your product passed quality checks. One step to publish." Always include the install command preview as identity reinforcement — seeing myclaude install @{username}/{slug} makes it real.
CRITICAL: Load references/ux-vocabulary.md for UX tier names and type names.
┌─────────────────────────────────────────────┐
│ ✦ {product_name} is ready to share! │
│ Quality: {ux_tier} {stars} │
│ │
│ {N} files packaged for {platform_count} │
│ platforms. Install command: │
│ │
│ myclaude install @{username}/{slug} │
│ │
│ {If value_score:} │
│ 💰 Suggested price: ${min}-${max} │
│ │
│ Next: /publish to go live │
└─────────────────────────────────────────────┘
Technical: {N} files, {size}, {checksum_short}...SAFETY INVARIANTS (D19 — critical constraints in high-attention zone)
[SOURCE: CC commit handling — never --amend, --no-verify, force push]
These are ABSOLUTE rules that cannot be overridden:
- Never overwrite .publish/ without confirmation — if .publish/ already exists, ask before replacing
- Never include secrets — re-run secrets scan from /validate Stage 2 on .publish/ contents AFTER stripping
- Never strip creator content — only strip WHY comments, _prefixed JSON keys, and .meta.yaml. All content the creator wrote must survive packaging intact
- Never modify workspace originals — all transformations happen on the .publish/ copy
- Verify frontmatter completeness — check
product-dna/{type}.yaml → frontmatter.required[]. Missing required fields in .publish/ → block with: "Required frontmatter field '{field}' missing. Add it to your product file, then re-run /package." [SOURCE: cc-platform-contract.md §2.1] - Compact Instructions preservation — if the workspace file has
## Compact Instructions, it MUST survive into .publish/. This section is NOT a WHY comment — it's functional product content read by Claude Code's compact system [SOURCE: compact/prompt.ts:133-143]
Anti-Patterns
- Packaging unvalidated product — Always check .meta.yaml state first.
- Modifying originals — Work on copies in .publish/. Never touch workspace/{slug}/ source files.
- Missing manifests — vault.yaml + plugin.json required. agentskills.yaml for compatible types. Triple distribution.
- Stale checksum — Recalculate if any file changes after initial staging.
- Silent failures — If WHY stripping breaks markdown structure, detect and report.
- Missing install artifacts — hooks without scripts/, statusline without settings-fragment.json. Step 6b catches these. If /validate passed but /package fails here, the DNA spec is missing a required file check.
- Shipping _comment keys in hooks.json — Step 2 strips
_-prefixed JSON keys. Verify after strip: no_comment,_why,_events_referencein .publish/ JSON files.
Compact Instructions
When context is compressed, preserve:
- Product slug, type, and packaging state
- Whether WHY stripping completed successfully
- Manifest generation status (vault.yaml, plugin.json, agentskills.yaml)
- Pricing intelligence results if computed
- Any blocking issues (secrets detected, frontmatter missing, JSON invalid)
- .publish/ directory path and file count
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

