Remove unused MCP servers, improve workflow execution by azalio · Pull Request #82 · azalio/map-framework

azalio · 2026-03-05T10:28:28Z

Summary

Remove context7 and claude-reviewer MCP servers from all configs, agents, templates, docs, tests, and scripts
Add resume detection to /map-plan — skip already-completed discovery/spec/plan steps on session restart
Auto-set batch mode in /map-efficient — skip CHOOSE_MODE, no more "step-by-step vs batch" prompt
Enforce parallel wave execution as REQUIRED (not optional) in /map-efficient
After Actor returns, forward to Monitor instead of manually debugging stale LSP diagnostics

Commits

ca8d749 — Remove context7 and claude-reviewer MCP servers (25 files, -324 lines)
fbd1a8a — Add resume detection to /map-plan
04172b2 — Auto-set batch mode, skip CHOOSE_MODE
fe46cf2 — Enforce parallel wave execution
29c8a0f — Don't debug after Actor, forward to Monitor

Test plan

pytest — 668 passed, 1 skipped, 0 failed
make sync-templates — templates in sync

Remove context7 and claude-reviewer from all MCP configurations, agent templates, docs, and tests. MAP now ships with only sequential-thinking and deepwiki as MCP servers. 25 files changed, -324/+67 lines. All 668 tests pass.

Check existing artifacts (findings, spec, task_plan, workflow_state) before starting each step. Skips already-completed phases when user restarts Claude mid-plan.

CHOOSE_MODE step is now auto-skipped by the orchestrator — execution mode is always batch. SUBTASK_APPROVAL (2.11) is also auto-skipped. resume_from_plan now jumps directly to INIT_STATE.

Add "always batch, always parallel" to Execution Rules (top-level). Mark Wave Computation as REQUIRED, not optional.

Add explicit instruction: after Actor returns, never manually debug stale LSP diagnostics. Always forward to Monitor for real compilation verification. Retry via Actor if Monitor reports valid=false.

Copilot

Pull request overview

This PR removes two unused MCP servers (context7 and claude-reviewer) from the MAP framework and makes several workflow execution improvements. The removal spans 25+ files including agent templates, command definitions, documentation, tests, scripts, and configuration files. The workflow improvements add resume detection to /map-plan, auto-set batch execution mode (eliminating the CHOOSE_MODE step), enforce parallel wave execution, and direct Actor output to Monitor instead of manually debugging stale LSP diagnostics.

Changes:

Remove all references to context7 and claude-reviewer MCP servers across agents, commands, configs, docs, tests, and scripts, reducing the server set to sequential-thinking and deepwiki
Add pre-flight resume detection to /map-plan and auto-set batch mode with CHOOSE_MODE auto-skip in /map-efficient orchestrator
Enforce parallel wave execution as required (not optional) and add Actor→Monitor forwarding guidance to prevent manual debugging of stale diagnostics

Reviewed changes

Copilot reviewed 30 out of 30 changed files in this pull request and generated 28 comments.

Show a summary per file

File	Description
`src/mapify_cli/__init__.py`	Remove context7/claude-reviewer from MCP server definitions, config creation, agent content generators, CLI help; remove "docs" MCP option
`tests/test_mapify_cli.py`	Update test assertions/comments from context7/claude-reviewer to deepwiki/sequential-thinking
`.map/scripts/map_orchestrator.py`	Auto-skip CHOOSE_MODE (always batch), update resume_from_plan to skip to INIT_STATE, remove validate_step check for execution_mode
`src/mapify_cli/templates/commands/map-plan.md` / `.claude/commands/map-plan.md`	Add pre-flight resume detection and per-step artifact checks
`src/mapify_cli/templates/commands/map-efficient.md` / `.claude/commands/map-efficient.md`	Auto-skip CHOOSE_MODE, enforce wave computation, add Actor→Monitor forwarding rule
`src/mapify_cli/templates/agents/.md` / `.claude/agents/.md`	Remove context7/claude-reviewer references from all agent templates (actor, monitor, predictor, evaluator, reflector, documentation-reviewer)
`src/mapify_cli/templates/commands/map-debug.md` / `.claude/commands/map-debug.md`	Remove context7 from debug MCP tools list
`docs/USAGE.md`, `docs/INSTALL.md`, `docs/CLI_COMMAND_REFERENCE.md`, `docs/ARCHITECTURE.md`	Update MCP server references, examples, and configuration docs
`.mcp.json.example`	Remove context7/claude-reviewer server configurations
`.claude-plugin/plugin.json`, `.claude-plugin/marketplace.json`, `.claude-plugin/PLUGIN.md`	Update MCP server requirements
`scripts/lint-agent-templates.py`	Remove context7 from MCP tool lint checks

Comments suppressed due to low confidence (1)

src/mapify_cli/templates/agents/monitor.md:527

Orphaned content after removing ### 3. mcp__context7__get-library-docs heading. Lines 524-527 still contain the body content for the removed context7 tool section. These should be either removed or restructured (e.g., integrated into the deepwiki section).

**Use When**: Code uses external libraries/frameworks
**Process**: `resolve-library-id` → `get-library-docs(library_id, topic)`
**Topics**: best-practices, security, error-handling, performance, deprecated-apis
**Rationale**: Current docs prevent deprecated APIs and missing security features

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-05T10:45:30Z

.claude/agents/documentation-reviewer.md

  → Use for library documentation verification
 ELSE:
  → Use Fetch to get raw documentation from official sources
-  → Log: "context7 unavailable, using direct fetch"



Same orphaned content issues as in the template: lines 110-112 have orphaned conditional body, and lines 423-425 have orphaned function parameters from the removed context7 call.

Copilot · 2026-03-05T10:45:31Z

.claude/agents/reflector.md


 2. Error involves library/framework misuse?
-   → context7 (resolve-library-id → get-library-docs)



Same orphaned content issue as in the template. Decision tree item "2. Error involves library/framework misuse?" at line 29 now has an empty body. Either remove this item or replace the removed context7 action with an appropriate alternative (e.g., → deepwiki for library documentation).

Suggested change

→ deepwiki for library documentation

Copilot · 2026-03-05T10:45:31Z

src/mapify_cli/templates/agents/actor.md

 ### Tool Selection Flowchart

 ```
 START → Using external library?


The flowchart's first decision "START → Using external library?" now only has "NO → Continue" — the "YES" path to context7 was removed. This makes the decision node meaningless since there's no "YES" action. Either remove the external library question entirely, or add deepwiki as the replacement action for "YES".

Suggested change

START → Using external library?

START → Using external library?

YES → deepwiki: read_wiki_structure → ask_question

Copilot · 2026-03-05T10:45:31Z

.claude/agents/actor.md

 ### Tool Selection Flowchart

 ```
 START → Using external library?


The flowchart's first decision "START → Using external library?" now only has "NO → Continue" — the "YES" path was removed. This makes the decision node meaningless. Either remove the question or add deepwiki as the replacement "YES" action.

Suggested change

START → Using external library?

START → Using external library?

YES → deepwiki: read_wiki_structure → ask_question

Copilot · 2026-03-05T10:45:31Z

src/mapify_cli/templates/agents/reflector.md

 - Use when: Library API misuse, verify usage patterns, recommend API changes
 - Process: resolve-library-id → get-library-docs with topic
 - Why: Ensure current APIs, avoid deprecated patterns


Orphaned bullet points after removing **mcp__context7__resolve-library-id + get-library-docs** heading. Lines 42-44 still contain the body content (use when, process, why) for the removed context7 tool section, but without any heading they appear to float under the sequential-thinking section. These lines should either be removed or refactored under a **mcp__deepwiki__...** heading if the guidance still applies.

Copilot · 2026-03-05T10:45:36Z

src/mapify_cli/templates/agents/documentation-reviewer.md

    topic="CRD installation and webhook requirements",
    tokens=3000


Orphaned code in the usage examples. After removing the mcp__context7__resolve_library_id and mcp__context7__get_library_docs function calls, lines 423-425 contain orphaned parameters (topic= and tokens= with a closing paren) that are no longer part of any function call. This is broken Python code inside a code block.

Suggested change

topic="CRD installation and webhook requirements",

tokens=3000

Fetch(

url="https://pypi.org/project/some-library/",

prompt="Verify CRD-related installation, webhooks, and required dependencies"

Copilot · 2026-03-05T10:45:36Z

.claude/agents/documentation-reviewer.md

    topic="CRD installation and webhook requirements",
    tokens=3000


Orphaned function parameters from the removed context7 call. Lines 423-425 contain topic= and tokens= parameters with a closing paren that are no longer part of any function call, resulting in broken code in the usage example.

Suggested change

topic="CRD installation and webhook requirements",

tokens=3000

Fetch(

url="https://openreports.io/docs/kubernetes/",

prompt="Verify CRD installation and webhook requirements"

Copilot · 2026-03-05T10:45:36Z

src/mapify_cli/templates/agents/reflector.md

  → Fall back to deepwiki for community documentation
  → Note: "Official docs unavailable, used community sources"


Orphaned fallback text. The IF context7 cannot resolve library: condition was removed, but its body (lines 355-356 "→ Fall back to deepwiki for community documentation" / "→ Note: ...") remains as dangling text inside the code block. This makes the edge case section incoherent — these lines should either be removed or restructured as a standalone fallback scenario.

Copilot · 2026-03-05T10:45:37Z

.claude/agents/reflector.md

  → Fall back to deepwiki for community documentation
  → Note: "Official docs unavailable, used community sources"


Orphaned fallback text. The IF context7 cannot resolve library: condition was removed, but its body remains as dangling text inside the code block.

Copilot · 2026-03-05T10:45:37Z

src/mapify_cli/templates/agents/evaluator.md

 **Use When**: Check consistency with past implementations
 **Rationale**: Maintain consistent standards (e.g., if past testability scored 8/10, use same criteria). Prevents score inflation/deflation.

-### 3. mcp__context7__get-library-docs
 **Use When**: Solution uses external libraries/frameworks
 **Process**: `resolve-library-id` → `get-library-docs(topics: best-practices, performance, security, testing)`
 **Rationale**: Libraries define quality standards (React testing, Django security). Validate solutions follow these.


Orphaned content after removing ### 2. mcp__claude-reviewer__get_review_history and ### 3. mcp__context7__get-library-docs headings. Lines 425-430 still contain the body text for both removed sections, floating without headings between the sequential-thinking section and ### 4. mcp__deepwiki__ask_question. These should either be removed or given proper headings.

azalio added 5 commits March 4, 2026 11:58

neutral: Remove context7 and claude-reviewer MCP servers

ca8d749

Remove context7 and claude-reviewer from all MCP configurations, agent templates, docs, and tests. MAP now ships with only sequential-thinking and deepwiki as MCP servers. 25 files changed, -324/+67 lines. All 668 tests pass.

neutral: Add resume detection to /map-plan

fbd1a8a

Check existing artifacts (findings, spec, task_plan, workflow_state) before starting each step. Skips already-completed phases when user restarts Claude mid-plan.

neutral: Auto-set batch mode in map-efficient, skip CHOOSE_MODE

04172b2

CHOOSE_MODE step is now auto-skipped by the orchestrator — execution mode is always batch. SUBTASK_APPROVAL (2.11) is also auto-skipped. resume_from_plan now jumps directly to INIT_STATE.

neutral: Enforce parallel wave execution in map-efficient

fe46cf2

Add "always batch, always parallel" to Execution Rules (top-level). Mark Wave Computation as REQUIRED, not optional.

neutral: Don't debug after Actor — forward to Monitor

29c8a0f

Add explicit instruction: after Actor returns, never manually debug stale LSP diagnostics. Always forward to Monitor for real compilation verification. Retry via Actor if Monitor reports valid=false.

Copilot AI review requested due to automatic review settings March 5, 2026 10:28

Copilot started reviewing on behalf of azalio March 5, 2026 10:29 View session

azalio merged commit f3af563 into main Mar 5, 2026
8 checks passed

Copilot AI reviewed Mar 5, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Remove unused MCP servers, improve workflow execution#82

Remove unused MCP servers, improve workflow execution#82
azalio merged 5 commits intomainfrom
neutral

azalio commented Mar 5, 2026

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Copilot AI Mar 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants


		2. Error involves library/framework misuse?
		→ context7 (resolve-library-id → get-library-docs)

	START → Using external library?
	START → Using external library?
	YES → deepwiki: read_wiki_structure → ask_question

		topic="CRD installation and webhook requirements",
		tokens=3000

		→ Fall back to deepwiki for community documentation
		→ Note: "Official docs unavailable, used community sources"

Conversation

azalio commented Mar 5, 2026

Summary

Commits

Test plan

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants