ask-claude

ask-claude is a skill for Claude Code from WebisityStudio/claude-codex-mcp-bridge. It costs 44 tokens per session (530 once invoked), scanned A, original, MIT.

A task handoff tool that starts a separate Claude Code worker chat for one bounded request. It can read a project, change files in an isolated Git worktree, and optionally open the finished chat in Claude's desktop app.

In plain words
What is it for?
Use it for separate implementation tasks, code inspections, or questions that another Claude Code session can handle. A Git worktree is a separate working copy of a repository.
Why use it?
It lets you delegate a focused coding task without mixing that work into the current conversation. Isolation helps keep file changes separate from the main checkout.

Skill for Claude Code

Written for Claude Code: user-invocable in frontmatter. Also seen: mentions Claude Code.

Good fit Use it for separate implementation tasks, code inspections, or questions that another Claude Code session can handle. A Git worktree is a separate working copy of a repository.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/webisitystudio/claude-codex-mcp-bridge/ask-claude
Install

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.

Any agent
npx skills add WebisityStudio/claude-codex-mcp-bridge --skill ask-claude
Clone the repo
git clone --depth 1 https://github.com/WebisityStudio/claude-codex-mcp-bridge

Made for: Claude Code.

Wrote 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.

agentmods badge for ask-claude

README.md
[![agentmods](https://agentmods.dev/badge/skills/webisitystudio/claude-codex-mcp-bridge/ask-claude/github.svg)](https://agentmods.dev/skills/webisitystudio/claude-codex-mcp-bridge/ask-claude)
Your own site
<a href="https://agentmods.dev/skills/webisitystudio/claude-codex-mcp-bridge/ask-claude"><img src="https://agentmods.dev/badge/skills/webisitystudio/claude-codex-mcp-bridge/ask-claude/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.

agentmods 80×15 button for ask-claude

Your own site · 80×15
<a href="https://agentmods.dev/skills/webisitystudio/claude-codex-mcp-bridge/ask-claude"><img src="https://agentmods.dev/badge/skills/webisitystudio/claude-codex-mcp-bridge/ask-claude.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 44 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 530 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00044 $0.00530
Opus 5.5 $0.00018 $0.00212
Sonnet 5.5 $0.00009 $0.00106
Haiku 4.5 $0.00004 $0.00053

Measured 6d ago against content hash 09412fdb6cce, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

ask-claude 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.

skills/ask-claude/SKILL.md · 25 lines

What it actually says

Ask Claude

Use the ask_claude MCP tool for the user's request. It starts a new Claude Code chat in the background; no mailbox or thread management is needed.

  1. Pass the user's task in request without diluting its acceptance criteria. The project defaults to this session's project; pass projectPath only for a different repository.
  2. Pick permissionMode by what the task needs:
    • plan or manual to inspect, review or answer a question. Nothing can change, and the real checkout is read.
    • acceptEdits (the default) to change files. The work happens in an isolated git worktree.
  3. Leave model and effort out unless the user asked for one or the task clearly needs it. The owner's policy caps what may be chosen and how many chats may be started. A refusal names the limit: ask for less or tell the user. Do not try to work around it.
  4. Set showInDesktop: true only when the user wants to see the chat in the Claude desktop app. The app switches to that chat when it opens.
  5. If the result warns about uncommitted changes and the task depends on them, run again with includeUncommitted: true.
  6. If the status is running_codex, the worker is still going: tell the user and carry on. The result arrives in your mailbox from bridge; you can also call bridge_orchestration_wait with the run ID. Never start a duplicate run.
  7. If the status is waiting_for_fable, answer the question and continue with bridge_continue_run using the returned run ID.
  8. Report changed files and real verification evidence. Compare filesChanged with the bridge's observedChanges, and pass on any warnings, such as tool calls the permission mode refused.
  9. Do not commit, push, merge, deploy, publish, send externally, alter credentials, delete data, or mutate production without explicit approval.

Claude workers run without MCP servers and need the Claude Code CLI signed in. If the tool reports a sign-in problem, tell the user to run claude auth login once; do not attempt it yourself.

Changes

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.

  1. 6d ago First seen · 25 lines · 44 tokens per session scan A 09412fdb6cce

Subscribe to this mod's changes

ask-claude is a skill published in the GitHub repository WebisityStudio/claude-codex-mcp-bridge (4 stars, last pushed 6d ago), licensed MIT. It adds 44 tokens to every session and 530 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.

Related

ask-claude compared with skills from other repositories

ask-claude next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
ask-claudeWebisityStudio/claude-codex-mcp-bridgeSkill46d agoMITA44Claude Code
agent-framework-py-releasemicrosoft/agent-frameworkSkill15ktodayMITA102Claude Code, Codex
coding-agents-farmgriddynamics/rosettaSkill3753d agoApache-2.0B32Claude Code
ghagents-squads/squads-cliSkill567d agoMITA0Claude Code, Codex
changelog-writer3338902669-ops/agent-orchestraSkill9424d agoApache-2.0A34Claude Code, Codex
release-notesspyrae/kronos-agent-osSkill6725d agoMITA19Claude Code, Codex
finishing-a-development-branchIbrahim-3d/orchestrator-supaconductorSkill3841mo agoAGPL-3.0A44Claude Code