/ask-gemini
Consult Gemini as an independent expert with deep web research. Sends a question or task to Gemini CLI and returns a research-backed response.
$ npx -y skills add PolyArch/humanize --skill ask-gemini --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/ask-gemini
Context preview
The summary Claude sees to decide when to auto-load this skill.
Consult Gemini as an independent expert with deep web research. Sends a question or task to Gemini CLI and returns a research-backed response.
SKILL.md
ask-gemini.SKILL.mdname: ask-gemini
description: Consult Gemini as an independent expert with deep web research. Sends a question or task to Gemini CLI and returns a research-backed response.
argument-hint: "[--gemini-model MODEL] [--gemini-timeout SECONDS] [question or task]"
allowed-tools: "Bash(${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh:*)"Ask Gemini
Send a question or task to Gemini and return a research-backed response. Gemini is always instructed to perform web research via Google Search, making this ideal for deep-research tasks that benefit from up-to-date internet information.
How to Use
Do not pass free-form user text to the shell unquoted. The question or task may contain spaces or shell metacharacters such as `(`, `)`, `;`, `#`, `*`, or `[`.
If the user only supplied a question or task, execute:
"${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh" "$ARGUMENTS"If the user supplied flags such as `--gemini-model` or `--gemini-timeout`, reconstruct the command so those flags remain separate shell arguments and the remaining free-form question is passed as one quoted final argument.
Example:
"${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh" --gemini-model gemini-2.5-pro "What are the latest Rust async runtime benchmarks?"Never run this unsafe form:
"${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh" $ARGUMENTSbecause the shell will re-parse the question text and can fail before `ask-gemini.sh` starts.
Interpreting Output
- The script outputs Gemini's response to **stdout** and status info to **stderr**
- Read the stdout output carefully and incorporate Gemini's response into your answer
- Gemini's responses are research-backed with web sources; relay source citations when available
- If the script exits with a non-zero code, report the error to the user
Error Handling
| Exit Code | Meaning | |-----------|---------| | 0 | Success - Gemini response is in stdout | | 1 | Validation error (missing gemini, empty question, invalid flags) | | 124 | Timeout - suggest using `--gemini-timeout` with a larger value | | Other | Gemini process error - report the exit code and any stderr output |
Notes
- The response is saved to `.humanize/skill/<timestamp>/output.md` for reference
- Default model is `gemini-3.1-pro-preview` with a 3600-second timeout
- Gemini is always instructed to perform Google Search for up-to-date information
Read more
name: ask-gemini
description: Consult Gemini as an independent expert with deep web research. Sends a question or task to Gemini CLI and returns a research-backed response.
argument-hint: "[--gemini-model MODEL] [--gemini-timeout SECONDS] [question or task]"
allowed-tools: "Bash(${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh:*)"Ask Gemini
Send a question or task to Gemini and return a research-backed response. Gemini is always instructed to perform web research via Google Search, making this ideal for deep-research tasks that benefit from up-to-date internet information.
How to Use
Do not pass free-form user text to the shell unquoted. The question or task may contain spaces or shell metacharacters such as `(`, `)`, `;`, `#`, `*`, or `[`.
If the user only supplied a question or task, execute:
"${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh" "$ARGUMENTS"If the user supplied flags such as `--gemini-model` or `--gemini-timeout`, reconstruct the command so those flags remain separate shell arguments and the remaining free-form question is passed as one quoted final argument.
Example:
"${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh" --gemini-model gemini-2.5-pro "What are the latest Rust async runtime benchmarks?"Never run this unsafe form:
"${CLAUDE_PLUGIN_ROOT}/scripts/ask-gemini.sh" $ARGUMENTSbecause the shell will re-parse the question text and can fail before `ask-gemini.sh` starts.
Interpreting Output
- The script outputs Gemini's response to **stdout** and status info to **stderr**
- Read the stdout output carefully and incorporate Gemini's response into your answer
- Gemini's responses are research-backed with web sources; relay source citations when available
- If the script exits with a non-zero code, report the error to the user
Error Handling
| Exit Code | Meaning | |-----------|---------| | 0 | Success - Gemini response is in stdout | | 1 | Validation error (missing gemini, empty question, invalid flags) | | 124 | Timeout - suggest using `--gemini-timeout` with a larger value | | Other | Gemini process error - report the exit code and any stderr output |
Notes
- The response is saved to `.humanize/skill/<timestamp>/output.md` for reference
- Default model is `gemini-3.1-pro-preview` with a 3600-second timeout
- Gemini is always instructed to perform Google Search for up-to-date information
Derived from the GAAC (GitHub-as-a-Context) project. A Claude Code plugin that provides iterative development with independent AI review. Build with confidence through continuous feedback loops.
Other skills on humanize.
- /ask-codex
Consult Codex as an independent expert. Sends a question or task to codex exec and returns the response.
Open skill - /humanize-gen-plan
Generate a structured implementation plan from a draft document. Validates input, checks relevance, analyzes for issues, and generates a complete plan.md with acceptance criteria.
Open skill - /humanize-refine-plan
Refine an annotated implementation plan into a comment-free plan and a QA ledger while preserving the gen-plan schema.
Open skill - /humanize-rlcr
Start RLCR (Ralph-Loop with Codex Review) on Codex using the native Stop hook.
Open skill - /humanize
Iterative development with AI review. Provides RLCR (Ralph-Loop with Codex Review) for implementation planning and code review loops.
Open skill

