Manus
Delegates deep-research and analysis tasks to the Manus AI agent via its REST API
Test report
- Verdict
- Works with setup
- Score
- Tested
- Jul 31, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · 92b1f6f
Cloned into a throwaway HOME=$(mktemp -d) and confirmed SKILL.md landed at ~/.claude/skills/manus/SKILL.md; frontmatter has name+description, the only referenced file (references/api.md) fetched HTTP 200, and a grep for curl|sh, base64 -d, eval, hardcoded /Users paths and injection phrasing found nothing. Trigger judging: I would load it for "Use Manus to research the best 4K monitors for Mac", "Delegate a competitive analysis of EV charging networks to Manus", "Check my Manus sessions and pull the results"; I would correctly skip "Refactor the auth middleware in this repo to use async/await"; but I judged "Research the EV charging market and summarize the key players" (no mention of Manus) as a likely false positive, since the description claims the skill for any "market analysis / competitive research" — 4/5. Output is unmeasured: I wrote a baseline delegation request and a skill-followed one (scope/timeframe/source/output-format constraints, agentProfile manus-1.6-max, poll + jq attachment extraction), but running the skill's actual first step returned {"code":16,"message":"missing authentication: require either API Key or Bearer Token"} — the research report itself needs a paid MANUS_API_KEY I do not have, so no real artifact exists to compare. Docs are honest and un-hyped, though neither SKILL.md nor the README mentions that tasks burn paid credits.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 5/5
- Triggers reliably 4/5
- Output vs. baseline 4/10
- Docs & honesty 4/5
What Manus does
Wraps the Manus AI agent REST API so Claude can hand off long-running research jobs — market analysis, product and stock comparisons, competitive landscaping, report generation — then poll the task until it finishes and extract the text output plus generated files. Triggers on phrasings like "use Manus", "delegate to Manus", "check my Manus sessions", and on multi-step research requests. Requires a MANUS_API_KEY from a Manus account; every operation is a curl call against api.manus.ai and consumes account credits.
How to install Manus
git clone --depth 1 https://github.com/sanjay3290/ai-skills.git /tmp/manus-src
mkdir -p ~/.claude/skills
cp -R /tmp/manus-src/skills/manus ~/.claude/skills/manus
# REQUIRED: a Manus account + API key from https://manus.im settings (paid credits are consumed per task)
# export MANUS_API_KEY=your-api-key # add to ~/.zshrc or ~/.bashrc
# Verified: POST https://api.manus.ai/v1/tasks returns 401 {"message":"missing authentication..."} without it.
# Optional deps: jq (the output/attachment extraction commands in SKILL.md pipe through jq).
# Plugin-marketplace alternative: /plugin marketplace add sanjay3290/ai-skills (repo ships .claude-plugin/marketplace.json)
# Skill files: SKILL.md + references/api.md (full endpoint reference, loaded on demand).
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Manus
-
/manusDelegates deep-research and analysis tasks to the Manus AI agent via its REST API
It also activates on plain-language prompts like these:
-
Send this competitor analysis to Manus for deep research -
Delegate the market comparison task to Manus -
Ask Manus to check on my running sessions
Frequently asked questions
- Is the Manus skill free?
- Yes. The skill itself is free from sanjay3290/ai-skills. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Manus work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 31, 2026. Verdict: Works with setup. Cloned into a throwaway HOME=$(mktemp -d) and confirmed SKILL.md landed at ~/.claude/skills/manus/SKILL.md; frontmatter has name+description, the only referenced file (references/api.md) fetched HTTP 200, and a grep for curl|sh, base64 -d, eval, hardcoded /Users paths and injection phrasing found nothing. Trigger judging: I would load it for "Use Manus to research the best 4K monitors for Mac", "Delegate a competitive analysis of EV charging networks to Manus", "Check my Manus sessions and pull the results"; I would correctly skip "Refactor the auth middleware in this repo to use async/await"; but I judged "Research the EV charging market and summarize the key players" (no mention of Manus) as a likely false positive, since the description claims the skill for any "market analysis / competitive research" — 4/5. Output is unmeasured: I wrote a baseline delegation request and a skill-followed one (scope/timeframe/source/output-format constraints, agentProfile manus-1.6-max, poll + jq attachment extraction), but running the skill's actual first step returned {"code":16,"message":"missing authentication: require either API Key or Bearer Token"} — the research report itself needs a paid MANUS_API_KEY I do not have, so no real artifact exists to compare. Docs are honest and un-hyped, though neither SKILL.md nor the README mentions that tasks burn paid credits.
- What is the Manus SkillProof Score?
- 6.8/10 — installs cleanly 5/5, triggers reliably 4/5, output vs. baseline 4/10, docs & honesty 4/5.
- How do I install Manus?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Manus with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.