AgentCanon MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
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
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| inspect_projectA | Detects repository structure, existing agent instruction files (AGENTS.md, CLAUDE.md, .cursor/rules), and AgentCanon status. |
| import_existing_rulesC | Bootstraps .agentcanon/rules.yaml from existing AGENTS.md, CLAUDE.md, or creates a starter configuration. |
| list_rulesC | Lists all canonical rules and metadata from .agentcanon/rules.yaml. |
| upsert_ruleC | Creates or updates a canonical rule inside .agentcanon/rules.yaml with Zod schema validation. |
| validate_rulesC | Validates canonical rules for schema errors, duplicate IDs, formatting conflicts, adapter warnings, and token budget. |
| preview_syncB | Computes deterministic agent projections, checks for manual drift/conflicts, and returns an approval-bound planId with file diffs. |
| apply_syncA | Applies a previously previewed sync plan after explicit user approval, verifying no files changed since preview. |
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 7 tools
Each tool targets a distinct phase of the rule-management lifecycle: inspection, import, listing, single-rule mutation, validation, plan generation, and plan application. The preview_sync/apply_sync pair is explicitly separated by the approval-bound planId contract, so there is no risk of confusing them. No two tools overlap in purpose.
All seven names use a consistent snake_case verb_noun pattern (inspect_project, list_rules, upsert_rule, validate_rules, preview_sync, apply_sync). Even the longer import_existing_rules follows the same verb-first convention.
Seven tools is a tight, well-scoped surface for a rule-canonicalization workflow, and each tool maps to a necessary step (read, import, list, write, check, plan, apply). Nothing appears redundant or missing from the count alone.
The core lifecycle from inspection through import, editing, validation, and guarded sync is fully covered. Minor gaps remain: there is no delete_rule or get_rule for fine-grained single-rule access, and no rollback/revert operation after apply_sync.