Skip to content
cccskills
FreePublished on GitHub

mcpb-bundle

How to migrate the tdmcp Codex Desktop bundle from legacy .dxt to .mcpb (MCP Bundle) and keep it building — covers the manifest schema (verify against the installed packer, never hardcode), the build-dxt.mjs packer/zip-fallback flow, the npm scripts, and the full .dxt→.mcpb reference sweep across docs and scripts. Use when packaging or migrating the tdmcp desktop bundle.

The English summary is being prepared. Showing the original description.

See how to install

Files included (1)

  • SKILL.md3.8 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

tdmcp Desktop bundle: .dxt → .mcpb

Anthropic renamed Desktop Extensions from DXT (.dxt, packer @anthropic-ai/dxt, manifest key dxt_version) to MCPB (.mcpb, packer @anthropic-ai/mcpb, manifest key manifest_version). Legacy .dxt still installs in Codex Desktop, but new directory submissions should ship .mcpb.

What this repo already has

scripts/build-dxt.mjs is already MCPB-aware: it tries @anthropic-ai/mcpb first, falls back to legacy @anthropic-ai/dxt, then to a system zip. The bundle stages manifest.json at the archive root + dist/, recipes/, td/, README.md, LICENSE, package.json, and a production-only node_modules. So the migration is mostly renaming the output + the references, not a rewrite.

dxt/manifest.json currently declares "manifest_version": "0.3" — already the modern MCPB key (not the legacy dxt_version), so the manifest is largely correct.

The migration, concretely

  1. Manifest — verify, don't guess. Before touching manifest_version, check what the installed packer accepts: npx --yes @anthropic-ai/mcpb --help (look for pack / validate). If it ships a validate command, run it against dxt/manifest.json and let it tell you. Only change manifest_version if validation demands it. A wrong value breaks install — this is why we verify rather than assume a number.
  2. Output filename → tdmcp.mcpb. In scripts/build-dxt.mjs: change outFile to tdmcp.mcpb and update the log lines (they say .dxt). Keep the packer-preference order and the zip fallback intact — both must still work. Optionally rename the script file to build-mcpb.mjs (update the npm script if you do).
  3. npm scripts (package.json): rename build:dxt → build:mcpb pointing at the script. Only keep a build:dxt alias if something external depends on the old name (grep first); otherwise replace it cleanly — no compatibility cruft. Check the version script too (it stages dxt/manifest.json).
  4. Reference sweep — change user-facing .dxt → .mcpb in: docs/guide/{install,troubleshooting,glossary}.md + their docs/pt/ mirrors, docs/DEPLOYMENT.md, docs/reference/cli.md, scripts/setup.mjs, README.md. Also grep .github/ for any release workflow that builds or uploads tdmcp.dxt / runs build:dxt.
  5. Release-asset URLs. Links to releases/latest/download/tdmcp.dxt point at a published asset. After this change the next release ships tdmcp.mcpb; update those links to .mcpb and record that a new release must be cut so the asset exists (the current v0.3.0 asset is still .dxt). Note this for the human in migration notes.

Preserve, don't erase

Where docs explain "one-click Desktop Extension", keep the explanation and add that .dxt still installs (legacy) while .mcpb is current. Don't delete info that helps users who already downloaded the .dxt.

Verify the build still works

npm run build          # populate dist/
npm run build:mcpb     # (renamed) → must emit tdmcp.mcpb
unzip -l tdmcp.mcpb     # manifest.json at root + dist/ present

The dxt/ directory name and dxt/manifest.json path can stay as-is (internal paths the script references) — renaming the directory is optional churn and not required for a valid .mcpb. If you rename it, update build-dxt.mjs, package.json's version script, and sync-manifest-version.mjs.

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Use when manage multiple local CLI agents via tmux sessions (start/stop/monitor/assign) with cron-friendly scheduling.

The English summary is being prepared. Showing the original description.

hybridlabor-api/aos6Updated Oct 16, 2026

Use when a hybrid memory system that provides persistent, searchable knowledge management for AI agents (Architecture, Patterns, Decisions).

The English summary is being prepared. Showing the original description.

hybridlabor-api/aos6Updated Oct 16, 2026

Reference for how BDB structures autonomous software engineering work — the seven-node dispatcher graph (Architect, TechLead, UI/UX, Engineering, Media/EventTech, Reviewer, Shipping) that /startcycle-graph actually runs. Use when you need the high-level lifecycle framing without inventing your own process.

The English summary is being prepared. Showing the original description.

hybridlabor-api/aos6Updated Oct 16, 2026

Tools are how AI agents interact with the world. A well-designed tool is the difference between an agent that works and one that hallucinates, fails silently, or costs 10x more tokens than necessary. This skill covers tool design from schema to error handling.

The English summary is being prepared. Showing the original description.

hybridlabor-api/aos6Updated Oct 16, 2026

Harness patterns for coding agents — memory, permissions, context engineering, delegation, skills, hooks, bootstrap.

The English summary is being prepared. Showing the original description.

hybridlabor-api/aos6Updated Oct 16, 2026

agenttrail: live map of a multi-agent build in the browser: which plan component is being worked on, by which agent or harness, what is done and what is stuck. Use when a multi-agent pipeline starts (/startcycle, /startcycle-graph, /teamwork-preview) or after a plan-canvas approve, when the user asks to see what the agents are doing ("live map", "build board", "who is running now"), or by yourself whenever a multi-agent run is under way — no user prompt needed.

The English summary is being prepared. Showing the original description.

hybridlabor-api/aos6Updated Oct 16, 2026

See all skills by hybridlabor-api

Report a problem with this skill