Tool Reference

architect-mcp exposes a small default core surface and an explicit advanced surface. The default public story is the agent work gate: clarify before edits, constrain the plan, review drift, and require verification evidence.

Historical V-labels remain in tool, script, test, and compatibility file identifiers. They are internal compatibility references, not public product framing.

Surfaces

• Core: the eight work-gate tools listed in the Core Work Gate section.
• Advanced: the core tools plus the compatibility, maturity, governance, repo-quality, stack-pack, MCP, and eval tools listed below.

Core Work Gate

The default public surface for clarifying work, constraining edits, reviewing drift, and checking final evidence.

Tool	Purpose
grill_me	Pressure-test the request before planning or editing.
create_pre_edit_contract	Turn clarified intent into an explicit edit contract.
review_build_plan	Review the implementation plan against the agreed contract.
review_proposed_file_plan	Check file paths and ownership before edits start.
review_repo_structure	Review repository structure from supplied file summaries.
review_implementation_against_contract	Compare changed files and verification against the pre-edit contract.
review_agent_final_response	Check the final response for evidence, receipts, honesty, and not-done disclosure.
review_agent_session	Combine intent, contract, drift, receipts, memory, and final-response checks into one session review.

Stack Packs And Standards

Stack-pack loading, source ingestion, promotion, standards comparison, and policy set tuning.

Tool	Purpose
list_stack_packs	List Stack Packs.
validate_stack_packs	Validate Stack Packs.
list_foundation_packs	List Foundation Packs.
validate_foundation_packs	Validate Foundation Packs.
stack_pack_expansion_strategy	Stack Pack Expansion Strategy.
discover_llms_sources	Discover Llms Sources.
fetch_llms_source	Fetch Llms Source.
ingest_llms_txt	Ingest Llms Txt.
list_ingested_llms_sources	List Ingested Llms Sources.
derive_stack_pack_from_llms_source	Derive Stack Pack From Llms Source.
propose_stack_pack_rules	Propose Stack Pack Rules.
review_stack_pack_candidate	Review Stack Pack Candidate.
promote_stack_pack_candidate	Promote Stack Pack Candidate.
promote_stack_pack_to_files	Promote Stack Pack To Files.
analyze_stack_pack_conflicts	Analyze Stack Pack Conflicts.
diff_stack_pack_versions	Diff Stack Pack Versions.
score_stack_packs	Score Stack Packs.
stack_pack_coverage_matrix	Stack Pack Coverage Matrix.
resolve_standards_profile	Resolve Standards Profile.
analyze_standards_conflicts	Analyze Standards Conflicts.
compare_standards_profiles	Compare Standards Profiles.
review_standards_refactor	Review Standards Refactor.
minimize_policy_set	Minimize Policy Set.
calibrate_rule_impact	Calibrate Rule Impact.

Repo Intake And Scaffolding

Repo profile selection, intake, artifact generation, scaffold planning, and repo-readiness checks.

Tool	Purpose
list_repo_profiles	List Repo Profiles.
get_repo_profile	Get Repo Profile.
infer_repo_layout	Infer Repo Layout.
forge_foundry_previews	Generate preview-only PR, issue, exception, no-op, and human-question artifacts from public-safe Foundry decisions without mutating repositories.
route_foundry_decisions	Route scored Foundry assessments into public-safe ledger entries without mutating repositories.
score_foundry_actionability	Score normalized Foundry evidence for PR-preview suitability, human review, exceptions, and no-op candidates.
normalize_foundry_evidence	Normalize findings, verification, repo constitution, coverage, and suppression candidates into public-safe Foundry evidence.
derive_repo_constitution	Derive repo instructions, PR templates, CI, release, package, and recent-PR style signals from supplied evidence.
derive_local_repo_constitution	Local-only scan that derives a public-safe repo constitution without mutating files.
start_project_intake	Start Project Intake.
grill_project_brief	Grill Project Brief.
continue_grill_me	Continue Grill Me.
generate_architecture_contract	Generate Architecture Contract.
generate_agent_instructions	Generate Agent Instructions.
generate_agents_md	Generate Agents Md.
generate_cursor_rules	Generate Cursor Rules.
generate_repo_artifacts	Generate Repo Artifacts.
generate_repo_scaffold_plan	Generate Repo Scaffold Plan.
review_local_workspace	Review Local Workspace.
create_review_baseline	Create Review Baseline.
validate_architecture_contract	Validate Architecture Contract.
validate_repo_artifacts	Validate Repo Artifacts.
self_review_architect_mcp	Self Review Architect Mcp.
diff_architecture_contracts	Diff Architecture Contracts.
mcp_readiness_report	Mcp Readiness Report.
score_repo_profile_fit	Score Repo Profile Fit.

Contract And Session Review

Ambiguity classification, implementation intent, contract lifecycle review, and session continuity checks.

Tool	Purpose
interpret_implementation_intent	Interpret Implementation Intent.
classify_ambiguity_risk	Classify Ambiguity Risk.
record_assumption	Record Assumption.
load_triggered_stack_guidance	Load Triggered Stack Guidance.
audit_work_gate_completeness	Report whether direct MCP workflows supplied complete, fresh, ordered work-gate evidence.
create_work_gate_sequence_receipt	Create a public-safe direct-client receipt for gate order, confirmed inputs, evidence, and freshness.
explain_review_findings	Explain Review Findings.
simulate_policy_gate	Simulate Policy Gate.
review_contract_lifecycle	Review Contract Lifecycle.
summarize_session_continuity	Summarize Session Continuity.
cluster_review_findings	Cluster Review Findings.
generate_local_report_artifact	Generate Local Report Artifact.

Memory, Skills, And MCP Integrations

Advisory memory, skill catalog review, MCP security review, MCP catalog recommendations, and guarded install plans.

Tool	Purpose
extract_harness_memory	Extract Harness Memory.
apply_harness_memory	Apply Harness Memory.
review_memory_relevance	Review Memory Relevance.
list_mcp_server_catalog	List the curated MCP server catalog.
recommend_mcp_servers	Recommend MCP servers only after provider and boundary checks.
create_mcp_install_plan	Create a dry-run MCP client configuration plan.
review_mcp_install_plan	Review an MCP install plan for unsafe config, secrets, and unpinned packages.
apply_mcp_install_plan	Local-only dry-run/apply tool for project-local MCP client config.
list_skill_catalog	List Skill Catalog.
recommend_skills_for_project	Recommend Skills For Project.
review_supplied_skills	Review Supplied Skills.
review_mcp_config_security	Review Mcp Config Security.
scan_mcp_config_files	Scan Mcp Config Files.
list_client_integration_recipes	List Client Integration Recipes.
audit_hosted_tool_policy	Audit Hosted Tool Policy.

Eval Harnesses And Repo Quality

Deterministic evals, generated repo quality checks, and regression coverage analysis.

Tool	Purpose
run_v3_eval_harness	Run V3 Eval Harness.
score_agent_artifacts	Score Agent Artifacts.
run_v5_eval_harness	Run V5 Eval Harness.
run_v6_eval_harness	Run V6 Eval Harness.
run_v7_eval_harness	Run V7 Eval Harness.
run_v8_eval_harness	Run V8 Eval Harness.
run_v9_eval_harness	Run V9 Eval Harness.
run_v10_eval_harness	Run V10 Eval Harness.
analyze_regression_coverage	Analyze Regression Coverage.
evaluate_scenario_acceptance	Evaluate Scenario Acceptance.
build_quality_requirements_profile	Build Quality Requirements Profile.
evaluate_repo_plan_quality	Evaluate Repo Plan Quality.
audit_generated_repo_quality	Audit Generated Repo Quality.
suggest_quality_followup_questions	Suggest Quality Followup Questions.
run_repo_quality_eval_scenarios	Run Repo Quality Eval Scenarios.
run_foundry_eval_corpus	Run deterministic offline Foundry eval scenarios over large and small repo summaries without network access or mutation.

Governance And Operating Model

Policy bundles, governance artifacts, orchestration planning, scenario routing, and productization boundaries.

Tool	Purpose
list_policy_bundles	List Policy Bundles.
validate_policy_bundles	Validate Policy Bundles.
preview_policy_bundle	Preview Policy Bundle.
review_pattern_card	Review Pattern Card.
preview_pattern_card	Preview Pattern Card.
create_architecture_strategy_map	Create Architecture Strategy Map.
preview_change_what_if	Preview Change What If.
diagnose_agent_behavior	Diagnose Agent Behavior.
draft_rule_candidate	Draft Rule Candidate.
compare_review_trends	Compare Review Trends.
render_governance_pack	Render Governance Pack.
select_review_playbook	Select Review Playbook.
review_playbook_conformance	Review Playbook Conformance.
check_agent_collaboration_plan	Check Agent Collaboration Plan.
run_failure_mode_drills	Run Failure Mode Drills.
review_documentation_intelligence	Review Documentation Intelligence.
select_local_orchestration_recipe	Select Local Orchestration Recipe.
normalize_mcp_result	Normalize Mcp Result.
plan_context_budget	Plan Context Budget.
route_evidence	Route Evidence.
create_local_dry_run_plan	Create Local Dry Run Plan.
review_tool_loop_quality	Review Tool Loop Quality.
get_v10_productization_blueprint	Get V10 Productization Blueprint.
create_v10_implementation_slice_plan	Create V10 Implementation Slice Plan.
plan_primer_dashboard	Plan Primer Dashboard.
validate_v10_productization_boundary	Validate V10 Productization Boundary.

Hosted Policy

Hosted mode exposes only hosted-safe tools. Local-only tools are excluded from hosted HTTP registration even when the advanced surface is requested.

Local-only tools:

• promote_stack_pack_to_files
• review_local_workspace
• derive_local_repo_constitution
• scan_mcp_config_files
• apply_mcp_install_plan

Future-adapter tools currently return stateless proposals. Durable storage waits for an explicit adapter:

• extract_harness_memory
• apply_harness_memory
• review_memory_relevance

Unknown tools classify as unknown and must never be treated as hosted-safe.