Install
$ agentstack add mcp-mikielek-consult ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Consult Skill
consult is an Agent Skills-compatible skill for asking another coding agent for an independent, advisory consultation with mutation-restricted defaults: review, brainstorming pass, debugging partner, or second opinion. A host agent can use it to involve installed backends including Gemini, OpenCode, Claude, Codex, and Pi without making the user think about the underlying command wrapper.
The canonical skill lives at:
.agents/skills/consult
See [.agents/skills/consult/SKILL.md](.agents/skills/consult/SKILL.md) for usage and safety guidance, and [.agents/skills/consult/INSTALL.md](.agents/skills/consult/INSTALL.md) for agent installation options.
How To Use It
Ask your coding agent for the consultation you want in normal language:
Ask Gemini to review this change critically and return only actionable risks.
Consult Pi on the API design. Be thorough, but avoid overengineering.
Have Gemini and Pi discuss this architecture for two rounds, then synthesize the tradeoffs.
Ask OpenCode for debugging hypotheses and rank the next checks by expected signal.
The host agent handles backend selection, prompt framing, session continuity, and claim verification. For multi-agent discussion, it can orchestrate rounds by consulting one backend, feeding the result to another, and then summarizing what survived local verification.
Direct Wrapper Usage
The shell wrapper is useful for maintainers, debugging, or manual use.
List available backends:
.agents/skills/consult/scripts/consult.sh --list
Run a consultation:
.agents/skills/consult/scripts/consult.sh --to gemini "Review this change for risks."
Use --prompt when it is clearer for scripts or when the prompt begins with -:
.agents/skills/consult/scripts/consult.sh --to gemini --prompt "Review this change for risks."
Preview the exact backend command:
.agents/skills/consult/scripts/consult.sh --to pi --dry-run "Check the API design."
Run deterministic wrapper checks:
.agents/skills/consult/tests/wrapper.sh
Notes
- Backend CLIs must be installed and authenticated separately.
- Consultations are advisory; verify claims locally before acting on them.
- The strongest read-only guarantee is the Codex backend's OS sandbox. Other backends use
approval, plan, or tool allowlist controls as documented in the skill references.
--jsonoutput shape is backend-specific; Gemini, Claude, and OpenCode use their CLI formats,
Codex emits JSONL events, and Pi intentionally rejects consult --json.
- The project is licensed under Apache-2.0. See
LICENSEandNOTICE.
Source & license
This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: mikielek
- Source: mikielek/consult
- License: Apache-2.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.