Handoff gemini
Delegate long-context analysis, research, and large-document summarization to Google Gemini CLI (1M-token context) via non-interactive `gemini -p`, preserving Claude context. Use when the user says 'use gemini'/'hand off to gemini' or a task needs 100K+ tokens of context.From its SKILL.md
npx -y skills add claude-world/director-mode-lite --skill handoff-geminiAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- runs commandsInstructs the agent to run 6 commands, including `npm install -g @google/gemini-cli` and 5 more.
SKILL.md
2.3 KB, 484 tokens by cl100k_base, as published. Nobody here has run it
Handoff to Gemini CLI
Delegate tasks to Google Gemini CLI to save Claude context.
When to Use Gemini
| Use Gemini For | Keep in Claude |
|---|---|
| Long document analysis | Complex coding |
| Multimodal tasks | Architecture decisions |
| Research and summarization | Multi-step workflows |
| Large file comprehension | Problem analysis |
Prerequisites
# Install Gemini CLI (Google)
npm install -g @google/gemini-cli
# Or via: https://github.com/google-gemini/gemini-cli
Handoff Process
1. Prepare Context
## Task for Gemini
**Goal**: [What needs to be done]
**Files**: [Which files to analyze]
**Output**: [What format to return]
2. Generate Command
Pass the prompt with -p (non-interactive). Reference files or directories inline with @<path>, or pipe content via stdin. There is no -f flag.
# Analyze a directory (@ reads it recursively)
gemini -p "Summarize the components in @src/"
# Research task (no file context)
gemini -p "Research best practices for [topic]"
# Document analysis (single file via @)
gemini -p "Extract key points from @docs/spec.md"
# Large or generated context via stdin
cat docs/spec.md | gemini -p "Extract key points"
Note: A
src/**/*.ts-style glob only expands if the shell has globstar enabled (shopt -s globstarin bash). Prefer a directory reference like@src/, which Gemini reads recursively without relying on shell globbing.
3. Provide Instructions
- Why Gemini is suitable
- Expected output
- How to continue workflow
Example
## Task: Research API Best Practices
**Command:**
gemini -p "Research REST API versioning strategies. Summarize pros/cons."
**After:**
- Review research
- Share relevant findings
- Implement chosen strategy
Benefits
- Token Savings: Research tasks don't consume Claude context
- Long Context: Handles very large documents
- Specialization: Use each AI's strengths
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.
Gives 0 of the 12 instructions most research analysis skills give in 484 tokens
Counted across 1,213 of the 2,113 authors here whose files we hold, read 2026-09-06
- Cite sources for every important claimin 47 of 1213, across 38 files
- Separate facts from inferences and recommendationsin 21 of 1213, across 12 files
- Write findings to a markdown filein 19 of 1213
- Label every insight with a confidence levelin 18 of 1213, across 8 files
- Read product marketing context before asking questionsin 18 of 1213, across 8 files
- Rank themes by frequency and intensityin 16 of 1213, across 6 files
- Establish research mode before proceedingin 16 of 1213, across 6 files
- Segment survey responses by customer tier or tenurein 16 of 1213, across 6 files
- Categorize support tickets before analyzingin 16 of 1213, across 6 files
- Weight research sources from the last twelve monthsin 16 of 1213, across 6 files
- Use at least five data points per segmentin 15 of 1213, across 5 files
- Extract verbatim quotes for all research findingsin 15 of 1213, across 5 files
Said here and by no other author read
- Use Gemini for long document analysis
- Use Gemini for research and summarization
- Use Gemini for large file comprehension
- Prepare task goal and files in markdown
- Pass prompts using the non-interactive flag
- Reference files or directories using the at symbol
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.