opencode-skills-collection 4.0.19 → 4.0.21

package/bundled-skills/docs/users/claude-code-skills.md

@@ -17,7 +17,7 @@ Configure AAS Core for Claude Code, describe the task and constraints, let Claud
 - It lets Claude search the verified local catalog without loading the full library into context.
 - It preserves Claude's exact selection without using metadata as an eligibility gate.
 - It keeps MCP discovery read-only and CLI changes approval-gated.
-- It includes 1,997+ skills instead of a narrow single-domain starter pack.
+- It includes 2,001+ skills instead of a narrow single-domain starter pack.
 - It supports the standard `.claude/skills/` path and the Claude Code plugin marketplace flow.
 - It also ships generated bundle plugins so teams can install focused packs like `Essentials` or `Security Developer` from the marketplace metadata.
 - It includes onboarding docs, bundles, and workflows so new users do not need to guess where to begin.

package/bundled-skills/docs/users/gemini-cli-skills.md

@@ -12,7 +12,7 @@ Install into the Gemini skills path, then ask Gemini to apply one skill at a tim
 
 - It installs directly into the expected Gemini skills path.
 - It includes both core software engineering skills and deeper agent/LLM-oriented skills.
-- It helps new users get started with bundles and workflows rather than forcing a cold start from 1,997+ files.
+- It helps new users get started with bundles and workflows rather than forcing a cold start from 2,001+ files.
 - It is useful whether you want a broad internal skill library or a single repo to test many workflows quickly.
 
 ## Install Gemini CLI Skills

package/bundled-skills/docs/users/kiro-integration.md

@@ -18,7 +18,7 @@ Kiro is AWS's agentic AI IDE that combines:
 
 Kiro's agentic capabilities are enhanced by skills that provide:
 
-- **Domain expertise** across 1,997+ specialized areas
+- **Domain expertise** across 2,001+ specialized areas
 - **Best practices** from Anthropic, OpenAI, Google, Microsoft, and AWS
 - **Workflow automation** for common development tasks
 - **AWS-specific patterns** for serverless, infrastructure, and cloud architecture

package/bundled-skills/docs/users/usage.md

@@ -39,7 +39,7 @@ If you came in through a **Claude Code** or **Codex** plugin instead of AAS Core
 
 When you ran `npx agentic-awesome-skills` or cloned the repository, you:
 
-✅ **Downloaded 1,997+ skill files** to your computer (default: `~/.agents/skills/`; or a custom path like `~/.agent/skills/` if you used `--path`)
+✅ **Downloaded 2,001+ skill files** to your computer (default: `~/.agents/skills/`; or a custom path like `~/.agent/skills/` if you used `--path`)
 ✅ **Made them available** to your AI assistant
 ❌ **Did NOT enable them all automatically** (they're just sitting there, waiting)
 
@@ -231,7 +231,7 @@ Let's actually use a skill right now. Follow these steps:
 
 ## Direct-install Step 5: Pick Skills Manually
 
-Don't try to use all 1,997+ skills at once. Here's a sensible approach:
+Don't try to use all 2,001+ skills at once. Here's a sensible approach:
 
 If you want a tool-specific starting point before choosing skills, use:
 
@@ -362,7 +362,7 @@ Usually no, but if your AI doesn't recognize a skill:
 
 ### "Can I load all skills into the model at once?"
 
-No. Even though you have 1,997+ skills installed locally, you should **not** concatenate every `SKILL.md` into a single system prompt or context block.
+No. Even though you have 2,001+ skills installed locally, you should **not** concatenate every `SKILL.md` into a single system prompt or context block.
 
 The intended pattern is:
 

package/bundled-skills/docs/users/visual-guide.md

@@ -40,7 +40,7 @@ agentic-awesome-skills/
 ├── 📄 CONTRIBUTING.md                  ← Contributor workflow
 ├── 📄 CATALOG.md                       ← Full generated catalog
 │
-├── 📁 skills/                          ← 1,997+ skills live here
+├── 📁 skills/                          ← 2,001+ skills live here
 │   │
 │   ├── 📁 brainstorming/
 │   │   └── 📄 SKILL.md                 ← Skill definition
@@ -53,7 +53,7 @@ agentic-awesome-skills/
 │   │   └── 📁 2d-games/
 │   │       └── 📄 SKILL.md             ← Nested skills also supported
 │   │
-│   └── ... (1,997+ total)
+│   └── ... (2,001+ total)
 │
 ├── 📁 apps/
 │   └── 📁 web-app/                     ← Interactive browser
@@ -106,7 +106,7 @@ agentic-awesome-skills/
 
 ```
                     ┌─────────────────────────┐
-                    │  1,997+ SKILLS          │
+                    │  2,001+ SKILLS          │
                     └────────────┬────────────┘
                                  │
         ┌────────────────────────┼────────────────────────┐
@@ -207,7 +207,7 @@ If you want a workspace-style manual install instead, cloning into `.agent/skill
 │   ├── 📁 brainstorming/                 │
 │   ├── 📁 stripe-integration/            │
 │   ├── 📁 react-best-practices/          │
-│   └── ... (1,997+ total)                │
+│   └── ... (2,001+ total)                │
 └─────────────────────────────────────────┘
 ```
 

package/bundled-skills/gh-attach/SKILL.md

@@ -0,0 +1,151 @@
+---
+name: gh-attach
+description: "Upload and download GitHub user-attachments (screenshots, PDFs, zips, videos) from the terminal; use when asked to attach or embed a file in a PR, issue, or comment, or download an attachment URL."
+category: developer-tools
+risk: critical
+source: community
+source_type: community
+source_repo: sudosubin/gh-attach
+date_added: "2026-08-01"
+author: sudosubin
+license: MIT
+license_source: "https://github.com/sudosubin/gh-attach/blob/main/LICENSE"
+tags:
+  - github
+  - attachments
+  - screenshots
+  - gh-extension
+  - cli
+tools:
+  - claude-code
+  - codex-cli
+  - cursor
+  - copilot
+plugin:
+  targets:
+    codex: blocked
+    claude: blocked
+  setup:
+    type: manual
+    summary: "Installs and runs a third-party gh extension that needs a GitHub user_session cookie or GH_ATTACH_SESSION_TOKEN."
+    docs: SKILL.md
+---
+
+# Upload and download GitHub user-attachments (gh-attach)
+
+GitHub has **no public API** for user-attachments. The web UI uses an internal
+endpoint that mints `github.com/user-attachments` URLs whose visibility follows the
+repository they belong to. [`gh-attach`](https://github.com/sudosubin/gh-attach)
+(MIT, sudosubin) replicates that drag-and-drop flow as a `gh` CLI extension, so an
+agent can upload a local file from the terminal, get a URL back, and later download
+an attachment URL to a file.
+
+## Overview
+
+This skill drives `gh-attach` to turn a local file (screenshot, image, PDF, zip,
+log, or video) into a hosted GitHub `user-attachments` URL, then embeds that URL
+into a pull request, issue, or comment. It also downloads an existing attachment
+URL back to a local file. GitHub auto-renders the URL as an image, video, or file
+wherever it is pasted, and the URL inherits the repository's visibility, so a
+private-repo upload stays private. It works against GitHub Cloud and GitHub
+Enterprise Server.
+
+## When to Use This Skill
+
+Use this skill when asked to:
+
+- "Attach a screenshot to the PR" or "add an image to the PR description"
+- "Attach this file (PDF, zip, log, video) to the issue or comment"
+- "Embed before/after screenshots" in a PR, issue, or README
+- "Download this GitHub attachment" from a `user-attachments` URL
+
+## How It Works
+
+### Step 1: Verify prerequisites
+
+```bash
+gh auth status                                   # gh installed and authenticated
+gh extension list | grep -q 'gh attach' \
+  || gh extension install sudosubin/gh-attach    # review/pin the extension source first
+```
+
+Uploads use a GitHub `user_session` browser cookie, **not** the `gh` token (that
+endpoint rejects tokens). By default `gh` must be authenticated so `gh-attach` can
+select the matching browser account (Chromium family, Firefox family, or Safari).
+If the wrong account is selected, add `--browser <name> --profile <name>`. For
+headless or CI use, set `GH_ATTACH_SESSION_TOKEN` to the bare `user_session` cookie
+value and treat it as a full account credential.
+
+### Step 2: Upload
+
+```bash
+# Use an absolute quoted path; -R is optional inside a repo working dir.
+URL=$(gh attach "/abs/path/screenshot.png" -R <owner>/<repo>)
+```
+
+`gh attach` prints the URL on one line to **stdout**. For GitHub Enterprise Server,
+use `-R host/owner/repo`. Capture the output; it is the embeddable reference.
+
+### Step 3: Embed into the PR / issue / comment
+
+```bash
+printf '## Screenshots\n\n%s\n' "$URL" \
+  | gh pr comment <pr> -R <owner>/<repo> --body-file -
+```
+
+Use `gh pr edit`, `gh issue comment`, or `gh issue edit` with `--body-file -` for
+other targets. Always pass `--body-file -` (not inline `--body`) so multi-line
+bodies and special characters cannot break shell quoting. GitHub auto-renders the
+URL, so paste it as-is.
+
+### Step 4: Download
+
+```bash
+# Specify the destination explicitly.
+gh attach download "$URL" -O "/abs/path/out.png"
+```
+
+Downloads of private attachments use the active `gh` token, with browser cookies as
+an authorization fallback.
+
+## Examples
+
+- **Attach a screenshot to PR #42:** upload the file, then append the URL under a
+  `## Screenshots` heading in the PR body with `gh pr edit ... --body-file -`.
+- **Embed before/after screenshots in a README:** upload both files, paste the two
+  URLs into the README at the relevant section.
+- **Download an attachment for review:** run `gh attach download "$URL" -O out.zip`
+  to fetch a `user-attachments` file locally.
+
+## Best Practices
+
+- Resolve globs to absolute paths first, and quote paths that contain spaces or
+  Unicode.
+- For display sizing, embed an HTML tag instead of the bare URL:
+  `<img width="800" src="$URL">`.
+- In CI, set `GH_ATTACH_SESSION_TOKEN` from a dedicated bot account.
+- `gh-attach` can upload multiple files concurrently and emit Markdown or JSON
+  output with jq-style filtering when you need to script around the result.
+
+## Limitations
+
+- **Session cookie required.** A `user_session` cookie grants full account access
+  (it is not scoped like a PAT), so treat it like a password and prefer a bot
+  account in CI.
+- **Write access to the target repo is required** to upload.
+- **Private-repo attachments stay private:** the `user-attachments` URL inherits
+  repo visibility, so an anonymous fetch on a private repo returns 404 or 403 by
+  design.
+- GitHub Cloud and GitHub Enterprise Server each decide which file extensions and
+  content types they accept.
+- The skill embeds the URL itself; `gh attach` only prints it.
+
+## Security & Safety Notes
+
+- The `user_session` cookie and `GH_ATTACH_SESSION_TOKEN` are full-account
+  credentials. Never print them, never paste them on a command line, and never
+  commit them. Prefer a dedicated bot account for headless or CI use.
+- Uploaded attachments are auto-rendered by GitHub, so only upload files you intend
+  to share with everyone who can view the target repository.
+- Confirm the destination `-R <owner>/<repo>` before uploading so an attachment is
+  not created against the wrong repository.

package/bundled-skills/talivia-agent-kit/SKILL.md

@@ -0,0 +1,139 @@
+---
+name: talivia-agent-kit
+description: "Set up and verify Talivia revenue analytics through MCP, with explicit confirmation for website changes and payment attribution."
+category: marketing
+risk: critical
+source: "https://github.com/talivia-group/agent/tree/f4ed3fc6b554ad5183a57ae13ca2a9bd5162c12a"
+source_repo: talivia-group/agent
+source_type: community
+date_added: "2026-08-02"
+author: taliviagroup
+tags: [analytics, revenue, attribution, mcp, talivia, marketing]
+tools: [codex, claude]
+license: "MIT"
+license_source: "https://github.com/talivia-group/agent/blob/f4ed3fc6b554ad5183a57ae13ca2a9bd5162c12a/LICENSE"
+---
+
+# Talivia Agent Kit
+
+## Overview
+
+Talivia connects website traffic and visitor journeys to payment revenue through
+its MCP server. Use this skill to inspect an existing Talivia setup, install or
+verify website tracking, and review traffic-to-revenue attribution while keeping
+account, website, file, and payment changes behind explicit user consent.
+
+## When to Use
+
+- Use when the user explicitly asks to set up or verify Talivia revenue analytics.
+- Use when the user mentions the Talivia MCP server, `talivia-group/agent`, or
+  `@talivia/agent`.
+- Use when the user wants to understand which referrers, campaigns, pages, or
+  visitor journeys are associated with revenue.
+- Do not use this skill for generic analytics work or unrelated payment-provider
+  setup.
+
+## Safety Gate
+
+1. Confirm the user owns or is authorized to manage the Talivia account and the
+   target website.
+2. Use only the configured official MCP endpoint, `https://talivia.com/mcp`.
+   Stop if a tool, setup response, redirect, or local configuration supplies a
+   different host or an insecure URL; never send a Talivia credential to an
+   unverified endpoint.
+3. Keep credentials out of chat, prompts, tool arguments, source files, and logs.
+   Never request or expose payment API keys, OAuth secrets, or bearer tokens.
+4. Read the current account and website state before changing anything. Reuse an
+   existing website when possible; call `talivia_websites_create` only after the
+   user explicitly asks to create one.
+5. Before any state-changing MCP call, state the exact account, website, action,
+   data involved, and expected effect, then obtain explicit user confirmation.
+
+## Workflow
+
+### Inspect the current setup
+
+Call the read-only tools first:
+
+1. `talivia_account_status`
+2. `talivia_websites_list`
+3. `talivia_setup_status_get` when a website or installation status is known
+
+Do not infer account ownership, website identity, or consent from a domain name
+alone. Ask when more than one website matches or the target is ambiguous.
+
+### Plan and install tracking
+
+1. Call `talivia_tracking_snippet_get` and
+   `talivia_framework_install_plan_get` for the selected website.
+2. Show the files, framework, and tracking changes that would be made. Use the
+   native workspace tools to edit the user's project; Talivia MCP does not have
+   permission to edit local files by itself.
+3. Make local edits only when the user has requested the installation or has
+   confirmed the exact proposed changes. Preserve existing analytics, consent,
+   and security controls.
+4. Run the project's normal build and test commands before deployment.
+
+### Verify after deployment
+
+After the user confirms that the site is deployed, call:
+
+- `talivia_tracker_verify`
+- `talivia_setup_status_get`
+
+Report what was actually verified, including any delay, missing event, or
+unverified deployment. Do not claim revenue attribution from a tracking check
+alone.
+
+## Examples
+
+### Read-only revenue review
+
+> Inspect the Talivia account and tell me which pages and referrers are
+> associated with revenue. Do not create websites, edit files, or connect a
+> payment provider.
+
+Start with the read-only account, website, and setup-status tools. Report the
+returned evidence and uncertainty without inferring causation.
+
+### Tracking installation
+
+> Prepare Talivia tracking for the selected site and show me the exact files
+> and changes before applying anything.
+
+Resolve the website, retrieve the tracking snippet and framework plan, present
+the proposed local diff, and wait for confirmation before writing or deploying.
+
+### Connect payment attribution
+
+1. Explain that payment attribution starts a browser-based authorization flow
+   and identify the Talivia account and website involved.
+2. Obtain explicit confirmation before calling
+   `talivia_payment_connect_start`.
+3. Send the user only to the secure URL returned by the official Talivia flow.
+   Do not ask the user to paste payment credentials or API keys into chat.
+4. Finish with `talivia_payment_status_get` and
+   `talivia_checkout_attribution_guide_get`, and clearly separate connected
+   status from verified revenue data.
+
+## Limitations
+
+- This skill does not establish legal authority, cookie consent, privacy
+  compliance, or payment-provider permissions for the user.
+- Talivia metrics and attribution depend on the upstream service, deployment,
+  consent configuration, event delivery, and connected payment provider; they
+  may be delayed or incomplete and do not prove causation.
+- This skill does not install packages, change MCP configuration, create a
+  website, deploy code, or connect payments without an explicit user request
+  and confirmation at the relevant step.
+- The upstream CLI and MCP server are external software. Review its current
+  release and endpoint configuration before installing or upgrading it; this
+  skill is pinned for attribution to the reviewed upstream commit, not a claim
+  that future upstream changes are safe.
+- Stop and ask for clarification when the account, website, endpoint, consent
+  state, requested file changes, or payment scope is ambiguous.
+
+## Source
+
+- Upstream repository: [talivia-group/agent](https://github.com/talivia-group/agent/tree/f4ed3fc6b554ad5183a57ae13ca2a9bd5162c12a)
+- Reviewed package version: `@talivia/agent@0.1.0`

package/bundled-skills/unified-ai-gateway/SKILL.md

@@ -0,0 +1,214 @@
+---
+name: unified-ai-gateway
+description: Operate and evaluate Unified AI System through eight governed MCP tools while preserving fake-provider, authorization, and evidence boundaries.
+category: ai-ml
+risk: critical
+source: https://github.com/happy520ai/unified-ai-system/tree/master/skills/unified-ai-gateway
+source_repo: happy520ai/unified-ai-system
+source_type: official
+date_added: "2026-08-01"
+author: happy520ai
+tags: [ai-gateway, codex, mcp, self-hosted, governance]
+tools: [codex]
+license: Apache-2.0
+license_source: https://github.com/happy520ai/unified-ai-system/blob/master/LICENSE
+---
+
+# Unified AI Gateway
+
+## Overview
+
+Use the official `unified-ai-system` MCP server to inspect and exercise a local
+AI gateway without provider credentials. This skill file provides operating
+guidance; it does not install the server or change Codex configuration by
+itself. The official Codex plugin bundles the MCP definition, while skill-only
+installations require the manual setup below.
+
+## Prerequisites And Setup
+
+1. Confirm that Codex CLI and Docker are installed and Docker is running.
+2. If the eight tools are already visible, skip setup and do not register a
+   duplicate server.
+3. Explain the first stage: it downloads the reviewed immutable linux/amd64
+   image into Docker's cache, inspects its metadata and layer history, creates
+   but never starts a temporary container, exports its root filesystem, removes
+   that temporary container, and writes an inspection inventory to a temporary
+   directory. Obtain explicit user approval for those download and inspection
+   changes only.
+4. After that first approval, pull the reviewed platform manifest and complete
+   the inspection. Do not execute the image or register it yet:
+
+```bash
+IMAGE='ghcr.io/happy520ai/unified-ai-system/mcp-server@sha256:cc17e923335f953631f59fb6a5ffcdce0e12e16c5abf362f1d28747452adadee'
+PLATFORM='linux/amd64'
+REVIEW_DIR="$(mktemp -d)"
+
+docker pull --platform "$PLATFORM" "$IMAGE"
+docker image inspect "$IMAGE" --format 'Id={{.Id}} OS={{.Os}} Architecture={{.Architecture}} User={{json .Config.User}} Entrypoint={{json .Config.Entrypoint}} Cmd={{json .Config.Cmd}} Labels={{json .Config.Labels}}'
+docker image history --no-trunc "$IMAGE" > "$REVIEW_DIR/image-history.txt"
+
+REVIEW_CONTAINER="$(docker create --platform "$PLATFORM" --pull never --entrypoint /bin/true "$IMAGE")"
+docker export --output "$REVIEW_DIR/rootfs.tar" "$REVIEW_CONTAINER"
+docker rm "$REVIEW_CONTAINER"
+
+tar -tf "$REVIEW_DIR/rootfs.tar" > "$REVIEW_DIR/rootfs-files.txt"
+mkdir -p "$REVIEW_DIR/rootfs"
+tar --same-permissions -xf "$REVIEW_DIR/rootfs.tar" -C "$REVIEW_DIR/rootfs"
+find "$REVIEW_DIR/rootfs/app" -type f -print > "$REVIEW_DIR/app-files.txt"
+find "$REVIEW_DIR/rootfs/app" \( -type l -o -type f -links +1 \) -exec ls -ld {} + > "$REVIEW_DIR/app-links.txt"
+find "$REVIEW_DIR/rootfs/app" -type f -name '*.node' -exec sha256sum {} + > "$REVIEW_DIR/native-binaries.sha256"
+find "$REVIEW_DIR/rootfs" -type f \( -perm -0100 -o -perm -0010 -o -perm -0001 \) -print > "$REVIEW_DIR/executable-files.txt"
+find "$REVIEW_DIR/rootfs" -type f \( -perm -4000 -o -perm -2000 \) -print > "$REVIEW_DIR/suid-sgid-files.txt"
+find "$REVIEW_DIR/rootfs/app" -type f \( -name '.env' -o -name '.env.*' -o -name '*.pem' -o -name '*.key' -o -name '*.p12' -o -name '*.pfx' -o -name 'id_rsa*' \) -print > "$REVIEW_DIR/credential-like-files.txt"
+find "$REVIEW_DIR/rootfs/app" -type f -name package.json -exec grep -nHE '"(preinstall|install|postinstall|prepare|prepack|postpack)"' {} + > "$REVIEW_DIR/lifecycle-hooks.txt"
+grep -RInE 'child_process|spawn\(|fetch\(|AI_GATEWAY_MCP_URL|process\.env|writeFile|appendFile|unlink|rm\(' "$REVIEW_DIR/rootfs/app/packages/mcp-server/src" "$REVIEW_DIR/rootfs/app/packages/shared-sdk/src" > "$REVIEW_DIR/runtime-sensitive-code.txt"
+```
+
+If `sha256sum` is unavailable, use the platform's SHA-256 utility and preserve
+the same report. Keep the review directory until the report is accepted; its
+deletion is another filesystem change and requires approval for the exact path.
+
+5. Read every generated inventory and report the inspection before proceeding.
+   Compare it with the versioned
+   [image content review](https://github.com/happy520ai/unified-ai-system/blob/2967617bb266c79992a6ed40e70bb59e67f661ca/docs/security/mcp-image-review-0.3.2.md).
+   Require linux/amd64 manifest digest
+   `sha256:cc17e923335f953631f59fb6a5ffcdce0e12e16c5abf362f1d28747452adadee`,
+   config digest
+   `sha256:c28166511ca9d95ab6f4c0c8ac0cbc2acbbfdac20f971d074ff6744d39e0248f`,
+   source `https://github.com/happy520ai/unified-ai-system`, revision
+   `541430d68fac6b35c512ea7d2df20fe45334e0a5`, version `0.3.2`, license
+   `Apache-2.0`, entrypoint `docker-entrypoint.sh`, and command
+   `node packages/mcp-server/src/index.js`.
+
+   Report these reviewed risks explicitly: the image uses the default root
+   user; includes Debian shell/package utilities and 11 base-image SUID/SGID
+   files; contains 549 internal pnpm links, two native Node binaries, and eight
+   lifecycle-hook declarations; and starts a child gateway with loopback HTTP.
+   The optional `AI_GATEWAY_MCP_URL` can make an HTTP or HTTPS connection only
+   when explicitly passed. The registered command below passes no host files,
+   environment variables, or ports and disables container networking. Stop on
+   any mismatch, unexpected link, credential-like file, native binary, hook,
+   privileged file, or sensitive-code behavior.
+6. Explain the second stage: it persists a Codex MCP configuration and permits
+   Codex to launch the inspected image in a later task. Obtain a separate
+   explicit approval for registration and activation; the download approval
+   does not carry over.
+7. After that second approval, register the reviewed platform digest with
+   pulling, container networking, Linux capabilities, and privilege escalation
+   disabled, then inspect the stored configuration:
+
+```bash
+codex mcp add unified-ai-system -- docker run --rm -i --pull never --platform linux/amd64 --network none --cap-drop ALL --security-opt no-new-privileges ghcr.io/happy520ai/unified-ai-system/mcp-server@sha256:cc17e923335f953631f59fb6a5ffcdce0e12e16c5abf362f1d28747452adadee
+codex mcp get unified-ai-system --json
+```
+
+8. Restart Codex or open a new task, then use `/mcp verbose` to confirm that all
+   eight tools are available. Remove the registration when it is no longer
+   wanted:
+
+```bash
+codex mcp remove unified-ai-system
+```
+
+Removing the registration does not remove the pulled image from Docker's
+cache. Treat image-cache deletion as a separate host-state change and obtain
+approval before doing it.
+
+## When to Use This Skill
+
+- Use when a user asks whether Unified AI System is healthy or ready.
+- Use when a user wants a credential-free gateway chat proof.
+- Use when a user asks about the gateway's knowledge, workflow, or workforce
+  surfaces.
+- Use when a user wants evidence from the bundled MCP tools rather than a claim
+  inferred from documentation or process exit codes.
+
+Do not use this skill for generic model comparisons, unrelated MCP servers, or
+deploying a production gateway.
+
+## Workflow
+
+1. Confirm that the `unified-ai-system` MCP tools are available in the current
+   task. If they are absent, follow the approved setup above and wait for a
+   restarted or new task.
+2. Call `gateway_health`, then `gateway_readiness`, before attempting chat.
+3. Select the narrowest additional tool that answers the request.
+4. Report returned provider, execution mode, readiness, and blockers exactly.
+5. Separate transport success from product, production-readiness, autonomy, or
+   AGI claims.
+
+## Tool Map
+
+- `gateway_health`: managed gateway status and provider mode
+- `gateway_readiness`: chat-path readiness and blockers
+- `gateway_chat`: deterministic credential-free chat proof
+- `knowledge_readiness`: knowledge subsystem readiness
+- `workflow_health`: workflow subsystem status
+- `workflow_actions`: available workflow actions
+- `workforce_health`: workforce subsystem status
+- `workforce_agents`: available workforce agents
+
+## Example
+
+```text
+User: Check whether the local gateway is ready, then prove chat works safely.
+
+Agent:
+1. Call gateway_health.
+2. Call gateway_readiness.
+3. Call gateway_chat only if both results prove fake-provider mode.
+4. Report provider, model, execution mode, response, and every blocker.
+```
+
+## Safety Boundaries
+
+- Keep the credential-free local fake provider as the default.
+- Never request, read, or transmit provider credentials through this skill.
+- Do not enable or call a real provider without explicit scoped authorization.
+- Treat MCP registration, image pulls, container creation, networking, and
+  teardown as host-state changes that require informed user approval.
+- Never substitute a mutable tag, the multi-platform index, or an unreviewed
+  platform manifest for the reviewed linux/amd64 digest. Keep download and
+  inspection approval separate from registration and activation approval.
+- Keep `--pull never` in the registered command. If the reviewed image is
+  absent from the local cache, fail closed and return to the first approval
+  stage.
+- Keep `--network none`, `--cap-drop ALL`, and
+  `--security-opt no-new-privileges` in the registered command.
+- Do not claim production readiness, L5 autonomy, or AGI from a healthy handshake.
+- Treat a zero exit code as transport evidence, not proof that readiness gates
+  passed.
+
+## Limitations
+
+- This skill file does not bundle the MCP server, Docker image, or Codex
+  configuration. It only operates tools supplied by the separately installed
+  official integration.
+- It does not deploy, benchmark, or certify the gateway for production use.
+- The credential-free chat tool proves only the deterministic local fake path.
+- It does not configure real providers or handle provider credentials.
+- The published MCP image requires Docker.
+- The reviewed `0.3.2` path is limited to linux/amd64. Do not activate an arm64
+  or other platform image without a separate content review.
+- The image runs as the container's default root user and bundles the gateway
+  source, package-manager tooling, native dependencies, and base-image
+  SUID/SGID files. The registered command drops capabilities, prevents new
+  privileges, disables networking, and leaves the image in Docker's cache.
+- Existing Codex tasks may not hot-load a newly installed MCP configuration.
+
+## Troubleshooting
+
+- If the tools are missing after approved registration, inspect
+  `codex mcp get unified-ai-system --json`, then restart Codex or start a new
+  task.
+- If readiness is blocked, report the returned blocker instead of retrying chat
+  blindly.
+- If the runtime might use a real provider, stop before chat and keep the
+  session read-only.
+
+## Additional Resources
+
+- [Unified AI System](https://github.com/happy520ai/unified-ai-system)
+- [60-second Codex MCP quickstart](https://github.com/happy520ai/unified-ai-system/blob/master/docs/codex-mcp-quickstart.md)
+- [MCP server guide](https://github.com/happy520ai/unified-ai-system/blob/master/packages/mcp-server/README.md)
+- [MCP image content review](https://github.com/happy520ai/unified-ai-system/blob/2967617bb266c79992a6ed40e70bb59e67f661ca/docs/security/mcp-image-review-0.3.2.md)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "opencode-skills-collection",
-  "version": "4.0.19",
+  "version": "4.0.21",
   "description": "OpenCode CLI plugin that automatically downloads and keeps skills up to date.",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",