CoAgent8
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| AGY_PATH | No | Backend executable for Antigravity (agy); overrides backends.<id>.executable | |
| CODEX_HOME | No | Backend configuration location for Codex; setup and cleanup support only the defaults | |
| CODEX_PATH | No | Backend executable for Codex; overrides backends.<id>.executable | |
| CLAUDE_PATH | No | Backend executable for Claude Code; overrides backends.<id>.executable | |
| coagent8_DIR | No | Storage root for sessions, history and caches (default ~/.coagent8) | |
| GEMINI_CLI_HOME | No | Backend configuration location for Antigravity; setup and cleanup support only the defaults | |
| coagent8_CONFIG | No | Configuration file path | |
| coagent8_HISTORY | No | `off` turns archive files off | |
| CLAUDE_CONFIG_DIR | No | Backend configuration location for Claude Code; setup and cleanup support only the defaults | |
| CODEX_CONFIG_PATH | No | Backend configuration location for Codex; setup and cleanup support only the defaults | |
| coagent8_PRICES_URL | No | Price list URL or local file instead of OpenRouter | |
| coagent8_QUOTA_CACHE | No | Quota cache file |
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 | {} |
| resources | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| consultA | Second opinion on a question, design, bug or plan. Returns the answer, then "CoAgent8:" lines: backend, model, seconds, session handle. |
| reviewA | Independent review of Git changes: findings, then "Verdict: READY" or "Verdict: BLOCKED (reason)". Read-only: runs no checks; an unchanged complete repeat cannot help. |
| doctorC | Models per backend: efforts, output prices, approval need; or backend health. |
| sessionC | List, page history or close sessions. |
| cancelC | Stop the running turn: interrupt if possible, then end the CLI. |
| issueC | Sanitized bug report with GitHub link. Be concise and actionable: reproduction steps, expected/actual behavior, impact, sanitized telemetry (tool/backend/model, CLI versions, errors, timing, scope/check evidence). Mark unknown facts; never invent them. Keep issueUrl verbatim; do not decode or re-escape it. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| Active Sessions | List of active CoAgent8 conversation sessions with metadata and turn state. |
TDQS
Scored across 6 tools
Each tool has a clearly distinct role: diagnostics, second-opinion consultation, Git review, session management, cancellation, and bug reporting. Consult vs. review is the closest overlap, but the descriptions sharply scope review to Git changes with a verdict while consult is for general questions, designs, bugs, or plans.
All names are lowercase single words, which is readable, but there is no consistent verb_noun or category pattern. The set mixes nouns and verbs (session, issue, doctor vs. consult, review, cancel), making the naming convention only partially predictable.
Six tools is well within a sensible range for this server's apparent consultation/review/session-management purpose. Each tool appears to earn its place without excessive fragmentation.
The surface covers diagnostics, consultation, review, session listing/history/closing, cancellation, and issue reporting, which is strong lifecycle coverage for the apparent domain. Minor gaps remain around explicitly starting/resuming sessions or follow-up interactions, but agents can likely work around them.