AgentStack
SKILL verified MIT Self-run

Route Gemini Context

skill-zacharticulatev-designer-pro-and-seo-route-gemini-context · by ZachArticulateV

Delegates large multi-file or whole-repo work to Gemini for architecture maps, refactor-impact analysis, and multi-document synthesis. Uses Gemini's large context window when the gemini CLI is connected; otherwise runs a scoped Claude pass over the most relevant files (narrower, lower-fidelity). Trigger when the user says "scan the whole repo", "long context", "all the files", "synthesize across…

No reviews yet
0 installs
7 views
0.0% view→install

Install

$ agentstack add skill-zacharticulatev-designer-pro-and-seo-route-gemini-context

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No 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.

Are you the author of Route Gemini Context? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

route-gemini-context

Family: routing Status: Stable

Purpose

The execution arm for tasks that need a context window bigger than Claude's: whole- repo architecture maps, refactor-impact analysis, multi-document synthesis, and large structured-data passes. Gemini returns a synthesized result that Claude then integrates.

Triggers

  • "scan the whole repo" / "long context" / "all the files" / "synthesize across these files"
  • "gemini analysis" / "map the architecture"
  • "trace dependencies" / "what changes if I refactor X"

Inputs

  • Task description; the file/directory scope; the specific question

Steps

  1. Check Gemini is available: gemini --version. If absent, tell the user and

fall back to a scoped Claude pass over the most relevant files (narrower, but honest about the limit).

  1. Scope the context — name the directories/files. For long media (video/PDF),

preprocess first (cap duration/pages) so you don't blow the context.

  1. Invoke Gemini with a focused ask and a required output format:

`` gemini -p "Map the architecture. Return a file:line list of where X happens. Cap 800 words." @./src ` Prefer the adjacent cc-gemini-plugin` (it packages directories) when available.

  1. Demand citations — file:line, page numbers, or timestamps — not a flat

summary, so the result is verifiable.

  1. Integrate the synthesis into the conversation; spot-check key claims against

the actual files before acting on them.

Outputs

  • Gemini's synthesized analysis (with file:line / page / timestamp citations)
  • A short note of what context was sent

Dependencies

  • gemini CLI, or the adjacent cc-gemini-plugin (optional — adds a context

window far larger than Claude's for whole-repo passes; free path: a scoped Claude pass over the most relevant files, narrower but honest about the limit).

Notes

Use when the task genuinely needs breadth. For a surgical lookup in a known file, use Claude directly — Gemini's strength is breadth, not precision. Verify before acting (recalled/long-context output can drift).

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.