zwilling
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| ZWILLING_ROOT | No | Project root. Defaults to $CLAUDE_PROJECT_DIR, otherwise the server's cwd. File paths are confined to it, agy runs in it, and the log lives in it. | |
| ZWILLING_TIMEOUT_MS | No | Kill agy after this many milliseconds and return a timeout error. | 120000 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| resources | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| ask_geminiA | Ask Gemini (via the Antigravity CLI) a question or hand it a task, optionally with files as context. Gemini runs without tool access: it only sees the prompt and the inlined files. Good for second opinions and reading many large files at once. |
| gemini_reviewA | Have Gemini review files and/or a git diff and list concrete issues (file:line, severity, suggested fix). Returns the written review plus structured findings. Give files, a diff, or both. |
| gemini_patchA | Have Gemini propose a code change as a unified diff. zwilling checks the paths it touches (same rules as attachments) and dry-runs it with |
| gemini_taskB | Run a named task template from .zwilling/tasks/.md in the project (e.g. explain, test-ideas, migration-plan), with optional extra instructions, files and diff. Gemini still has no tool access. No task templates found in this project yet. |
| list_gemini_modelsA | List the model IDs accepted by the |
| read_gemini_logA | Return entries of the shared Gemini log (docs/ai/GEMINI_LOG.md), to recall earlier Gemini answers. Filter by id, tool or a search text; returns the last N matches. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| gemini-log | The shared Gemini log (docs/ai/GEMINI_LOG.md). |
TDQS
Scored across 6 tools
The four Gemini-calling tools have reasonably distinct framing: ask_gemini for open questions, gemini_review for issue-finding, gemini_patch for diffs, gemini_task for templated runs. There is still conceptual overlap (a well-worded ask_gemini or a review-style task template can approximate gemini_review), but the descriptions steer selection adequately. The two supporting tools (list_gemini_models, read_gemini_log) are clearly separate.
All names are snake_case, but three different patterns coexist: verb_gemini_noun (list_gemini_models, read_gemini_log), verb_noun (ask_gemini), and gemini_noun (gemini_review, gemini_patch, gemini_task). The split between 'ask_gemini' and 'gemini_review' is especially arbitrary. Readable, but no predictable rule an agent can rely on.
Six tools is well within the sweet spot and each maps to a distinct workflow: model discovery, log recall, open Q&A, review, patch proposal, templated tasks. Nothing looks like filler or a redundant alias.
Core lifecycle for a Gemini-assist wrapper is covered: discover models, ask, review, patch, run templates, and recall history. The main gap is task-template discoverability — gemini_task requires knowing .zwilling/tasks/<task>.md names, and there is no list_task_templates tool (the description even notes no templates were found). Applying a patch is deliberately excluded and documented, so that is not a real gap.