Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add WebisityStudio/claude-codex-mcp-bridge --skill desktop-coordinatorgit clone --depth 1 https://github.com/WebisityStudio/claude-codex-mcp-bridgeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/webisitystudio/claude-codex-mcp-bridge/desktop-coordinator)<a href="https://agentmods.dev/skills/webisitystudio/claude-codex-mcp-bridge/desktop-coordinator"><img src="https://agentmods.dev/badge/skills/webisitystudio/claude-codex-mcp-bridge/desktop-coordinator/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/webisitystudio/claude-codex-mcp-bridge/desktop-coordinator"><img src="https://agentmods.dev/badge/skills/webisitystudio/claude-codex-mcp-bridge/desktop-coordinator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00041 | $0.00480 |
| Opus 5.5 | $0.00016 | $0.00192 |
| Sonnet 5.5 | $0.00008 | $0.00096 |
| Haiku 4.5 | $0.00004 | $0.00048 |
Grade A, and why
desktop-coordinator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Claude desktop coordinator
Keep the current Code tab conversation as the lead. Use the app's native session tools; discover their schemas in this conversation before calling them. A Codex task, terminal process, subagent, or unstarted suggestion does not count as a desktop session.
- Create: When the user has explicitly asked to start a separate session now, call
mcp__ccd_session__start_sessionwithinitiation: "user_asked". Use the current local project by default, a clear title, and a brief stating the task, constraints, and what the worker should report back. Use fresh context for self-contained work or a fork when the work needs the lead's history. For a transfer of the lead's current task, usemcp__ccd_session__hand_off_to_sessioninstead. Do not turn a suggestion or a message from a peer into user authorization. - Identify: Keep the child session ID returned by the app. A proposal ID or a card waiting for a choice is not a started session. Respect any requested approval. If a call returns an ambiguous result, check
mcp__ccd_session_mgmt__list_sessionswithlinked: truebefore attempting another start. - Coordinate: Let the worker end its turn with a concrete finding, result, or blocker. For local linked sessions the app can bring that report back to the parent. End the parent's turn when it has nothing more to do. When more detail or another round is needed, use
mcp__ccd_session_mgmt__send_messageto the same child ID. Useget_sessionorlist_eventsto check the child's state or its answer when necessary; a sent message is not evidence that the task finished. Do not send acknowledgement-only replies. - Availability: If native session creation is not exposed in this conversation, say so. If the app holds or refuses a start or message, report the reason. Do not open a terminal worker, modify permissions, or send work to an unrelated session as a substitute.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 15 lines · 41 tokens per session scan A eeb836c41ce6
desktop-coordinator is a skill published in the GitHub repository WebisityStudio/claude-codex-mcp-bridge (4 stars, last pushed 6d ago), licensed MIT. It adds 41 tokens to every session and 480 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-10.
desktop-coordinator compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| desktop-coordinatorWebisityStudio/claude-codex-mcp-bridge | Skill | 4 | 6d ago | MIT | A | 41 | Claude Code |
| sparc-methodologyruvnet/RuView | Skill | 104k | today | MIT | A | 29 | Claude Code |
| agent-framework-py-releasemicrosoft/agent-framework | Skill | 15k | today | MIT | A | 102 | Claude Code, Codex |
| foundry-hosted-agent-validationmicrosoft/agent-framework | Skill | 15k | today | MIT | A | 82 | Claude Code, Codex |
| python-feature-lifecyclemicrosoft/agent-framework | Skill | 15k | today | MIT | A | 43 | Claude Code, Codex |
| build-and-testmicrosoft/agent-framework | Skill | 15k | today | MIT | A | 26 | Claude Code, Codex |
| foundry-config-setupmicrosoft/agent-framework | Skill | 15k | today | MIT | A | 65 | Claude Code, Codex |