Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault

No arguments

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

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}

Tools

Functions exposed to the LLM to take actions

NameDescription
project_queryA

Where is X / what handles Y: ranked code locations as text with numbered source lines. Hits are leads, not claims.

index_updateA

Re-index after editing files (a never-scanned folder: the first scan); claims on changed files become stale.

analyzeA

Start here for a why, when, which-commit or what-does-it-affect question: a verdict, answer claims and leads with file:line and commit hashes. Leads are what to read next, not claims; confirm one with your own git or grep. needs_clarification is normal: ask the user.

code_checkA

Python, Java, Kotlin, TS/JS imports; other languages: not_checked. For code you wrote or edited: do the modules, names, methods and arguments it uses exist? No arguments: the changes against HEAD. Each site: exists | absent (nearest names) | unknown | not_installed | guarded.

run_toolA

Run one more Verinoda tool: locate {text}, coupled {files}: files an issue touches, files that change together; node_inspect {name}: a symbol's definition and edges; relation_trace {source, target}: call paths between symbols; map_view {view, targets?}: architecture views (impact, hierarchy, ...); claim_list {status?}, claim_inspect {claim_id}, evidence_inspect {evidence_id}: claims, evidence; change_review {targets?, change?, since_last?}: what an edit touches; history_search {text|symbol|message|base}: when text came or went, commits; tq {questions}: typed questions, many per call.

Prompts

Interactive templates invoked by user choice

NameDescription
reviewReview the working-tree change (or the change since base): what it touches, names that do not exist, decision records, dependents.
onboardingGet to know the project (or one topic in it): structure, dependencies, tests, then how it works as claims with evidence.
debugDebug a failure: locate the code the symptom names, ask why, then record every fix attempt against one repro.
pre_mergeCheck a branch before merging into base: commits and files it adds, what they touch, names, dependencies, decision records, cycles.

Resources

Contextual data attached and managed by the client

NameDescription

No resources

TDQS

A3.6/5.0

Scored across 5 tools

Disambiguation3/5

project_query and analyze both return file:line leads and their boundary ('where is X' vs 'why/when/which-commit') is stated but easy to blur in practice. Worse, run_tool bundles locate, change_review, tq, map_view, and history_search, which substantially duplicate what project_query and analyze already do, so an agent must guess whether to call the top-level tool or the meta-tool.

Naming Consistency3/5

project_query and index_update follow noun_verb, code_check is noun_verb but reads oddly, run_tool is verb_noun, and analyze is a bare verb. Readable individually, but no single predictable pattern, so the set feels ad hoc.

Tool Count3/5

Five top-level tools is a sensible count for a code-intelligence server, but run_tool secretly exposes ~11 more operations (node_inspect, relation_trace, claim_list, history_search, tq, etc.). That makes the real surface ~16 tools funneled through one dispatch entry point, hurting discoverability more than a flat tool list would.

Completeness4/5

Coverage is strong for the domain: search (project_query/locate), indexing (index_update), reasoning over history and architecture (analyze, map_view, history_search, relation_trace), verification (code_check), and a claim/evidence model. Minor gaps exist around inspecting raw file contents or index health/status, but core workflows have no dead ends.

Maintenance

ActivityMaintained
ResponsivenessNo recent community issues