local-coder-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| OLLAMA_BASE_URL | No | Local Ollama server | http://127.0.0.1:11434 |
| LOCAL_CODER_MODEL | No | Local executor model | qwen2.5-coder:14b |
| LOCAL_CODER_TIMEOUT_MS | No | Ollama request timeout | 180000 |
| LOCAL_CODER_MAX_FILE_BYTES | No | Maximum size of one supplied/generated file | 120000 |
| LOCAL_CODER_TELEMETRY_PATH | No | Telemetry JSONL path | ~/.local-coder-mcp/telemetry.jsonl |
| LOCAL_CODER_ALLOWED_COMMANDS | No | Validation executable allowlist | npm,pnpm,yarn,bun |
| LOCAL_CODER_MAX_CONTEXT_BYTES | No | Maximum combined repository file context | 600000 |
| LOCAL_CODER_TELEMETRY_ENABLED | No | Enable aggregate local telemetry | true |
| LOCAL_CODER_VALIDATION_TIMEOUT_MS | No | Timeout for each validation command | 180000 |
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
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| local_coder_healthA | Check Ollama connectivity plus fast/strong adaptive model availability and resource settings. |
| classify_local_code_taskB | Classify as deterministic, local, local-supervised, or Claude. local-supervised means Claude already resolved a sensitive auth/credential/permission/security decision and only bounded implementation remains; full-diff Claude review is mandatory. |
| discover_local_workspaceA | List a bounded safe workspace view and root package scripts without following symlinks. |
| search_local_workspaceB | Literal bounded text/code search inside a safe workspace. |
| delegate_code_taskB | Ask the fast local model for bounded code/analysis text without modifying repository files. |
| execute_local_code_taskC | Compatibility full-result executor. Prefer execute_local_code_task_compact, which includes routing preflight, adaptive fast-to-strong retry, and supervised-sensitive review enforcement. |
| execute_local_code_planC | Compatibility full-result orchestrator. Supports local-supervised subtasks and adaptive fast-to-strong retry; prefer the compact orchestrator for enforced review behavior. |
| prepare_local_contextA | Build a compact evidence-first context capsule for a coding task using a persistent local repository index. Prefer this before broad Claude file exploration; verify cited file:line evidence for architectural or risky decisions. |
| execute_local_code_task_compactB | Preferred bounded local executor. It preflights routing itself. Sensitive auth/credential/permission/security implementation may run as local-supervised only after Claude explicitly resolves the sensitive decision; supervised runs force full-diff Claude review. |
| execute_local_code_plan_compactB | Preferred large-feature orchestrator. Claude owns planning/decomposition. Already-resolved sensitive subtasks can run as local-supervised; any such task forces full aggregate-diff review by Claude. |
| get_local_runA | Fetch a stored local execution result only when Claude needs more detail. Prefer summary first; request diff, validation, or full incrementally with offset/maxChars instead of loading a large result into context at once. |
| local_coder_telemetryB | Aggregate privacy-preserving routing/execution/orchestration telemetry including exact per-model Ollama inference usage. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 12 tools
Multiple execution tools overlap heavily: delegate_code_task, execute_local_code_task, execute_local_code_task_compact, execute_local_code_plan, and execute_local_code_plan_compact all route work to a local model, and the differences between compatibility and preferred variants are not crisp. Other tools like health, search, classify, context, and telemetry are distinct.
Most tools use snake_case verb_noun patterns like execute_local_code_task and get_local_run, but local_coder_health and local_coder_telemetry break the verb pattern and mix 'coder' with the 'code' used elsewhere. The compact/compat suffixes are consistent but add clutter.
Twelve tools is a reasonable count for an Ollama-backed coding assistant, and most cover distinct functions. However, several are compatibility or preferred duplicates that could be consolidated, making the set slightly less clean than ideal.
The set covers health, routing classification, workspace discovery/search, context preparation, task/plan execution, result retrieval, and telemetry, so the main workflows have no dead ends. Gaps like cancellation or listing of stored runs are minor and workaroundable.