mapify-cli 3.13.1__tar.gz → 3.14.0__tar.gz

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mapify-cli
-Version: 3.13.1
+Version: 3.14.0
 Summary: MAP Framework installer - Modular Agentic Planner for Claude Code
 Project-URL: Homepage, https://github.com/azalio/map-framework
 Project-URL: Repository, https://github.com/azalio/map-framework.git
@@ -201,7 +201,7 @@ These flows maintain branch-scoped artifacts under `.map/<branch>/` — `bluepri
 - **Daily-driver speed** — optimized for repeated use, not occasional demo workflows. Structured enough to prevent chaos, lightweight enough to keep token and time cost under control.
 - **Reviewable diffs** — `/map-plan` and `/map-efficient` require per-subtask size, concern, and constraint metadata, then validate `blueprint.json` *before* implementation, so oversized or mixed-concern plans fail early instead of surprising reviewers later.
 - **Gates that check the plan, not vibes** — `/map-check` and `/map-review` validate against the spec, tests, and diff instead of asking whether code "looks fine".
-- **Clean-room review** — `/map-review` auto-bundles spec, plan, tests, verification, and coverage evidence into a single durable input (`.map/<branch>/review-bundle.json`); `--detached` opens a read-only worktree for inspection without touching your branch.
+- **Clean-room review** — `/map-review` auto-bundles spec, plan, tests, verification, and coverage evidence into a single durable input (`.map/<branch>/review-bundle.json`); `--detached` opens a read-only worktree for inspection without touching your branch. With `minimality` enabled, review also runs an advisory what-to-delete lens for over-engineering cuts.
 - **Project memory** — `/map-learn` turns hard-won fixes and gotchas into reusable context, so the next session doesn't relearn them.
 
 <details>
@@ -249,7 +249,7 @@ MAP is inspired by the [MAP cognitive architecture](https://github.com/Shanka123
 minimality: lite   # new installs default to lite; existing repos without the key stay off
 ```
 
-Allowed values: `off`, `lite`, `full`, `ultra`. Phase 1 ships conservative `lite` defaults: Actor prefers the fewest moving parts, Monitor blocks scope drift only when it affects required behavior, and Evaluator scores simplicity without letting it hide missing required work.
+Allowed values: `off`, `lite`, `full`, `ultra`. Phase 1 ships conservative `lite` defaults: Actor prefers the fewest moving parts, Monitor blocks scope drift only when it affects required behavior, and Evaluator scores simplicity without letting it hide missing required work. `/map-review` also adds an advisory what-to-delete lens when minimality is not `off`; its `net: -N` estimate is informational, not a gate.
 
 **Context-compression policy** (controls the `/compact` nudge; default `never` — opt-in):
 

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/README.md

@@ -157,7 +157,7 @@ These flows maintain branch-scoped artifacts under `.map/<branch>/` — `bluepri
 - **Daily-driver speed** — optimized for repeated use, not occasional demo workflows. Structured enough to prevent chaos, lightweight enough to keep token and time cost under control.
 - **Reviewable diffs** — `/map-plan` and `/map-efficient` require per-subtask size, concern, and constraint metadata, then validate `blueprint.json` *before* implementation, so oversized or mixed-concern plans fail early instead of surprising reviewers later.
 - **Gates that check the plan, not vibes** — `/map-check` and `/map-review` validate against the spec, tests, and diff instead of asking whether code "looks fine".
-- **Clean-room review** — `/map-review` auto-bundles spec, plan, tests, verification, and coverage evidence into a single durable input (`.map/<branch>/review-bundle.json`); `--detached` opens a read-only worktree for inspection without touching your branch.
+- **Clean-room review** — `/map-review` auto-bundles spec, plan, tests, verification, and coverage evidence into a single durable input (`.map/<branch>/review-bundle.json`); `--detached` opens a read-only worktree for inspection without touching your branch. With `minimality` enabled, review also runs an advisory what-to-delete lens for over-engineering cuts.
 - **Project memory** — `/map-learn` turns hard-won fixes and gotchas into reusable context, so the next session doesn't relearn them.
 
 <details>
@@ -205,7 +205,7 @@ MAP is inspired by the [MAP cognitive architecture](https://github.com/Shanka123
 minimality: lite   # new installs default to lite; existing repos without the key stay off
 ```
 
-Allowed values: `off`, `lite`, `full`, `ultra`. Phase 1 ships conservative `lite` defaults: Actor prefers the fewest moving parts, Monitor blocks scope drift only when it affects required behavior, and Evaluator scores simplicity without letting it hide missing required work.
+Allowed values: `off`, `lite`, `full`, `ultra`. Phase 1 ships conservative `lite` defaults: Actor prefers the fewest moving parts, Monitor blocks scope drift only when it affects required behavior, and Evaluator scores simplicity without letting it hide missing required work. `/map-review` also adds an advisory what-to-delete lens when minimality is not `off`; its `net: -N` estimate is informational, not a gate.
 
 **Context-compression policy** (controls the `/compact` nudge; default `never` — opt-in):
 

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mapify-cli"
-version = "3.13.1"
+version = "3.14.0"
 description = "MAP Framework installer - Modular Agentic Planner for Claude Code"
 authors = [{ name = "MAP Framework Contributors" }]
 readme = "README.md"

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/src/mapify_cli/__init__.py

@@ -23,7 +23,7 @@ Or install globally:
     mapify check
 """
 
-__version__ = "3.13.1"
+__version__ = "3.14.0"
 
 import os
 import subprocess

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/src/mapify_cli/templates/hooks/safety-guardrails.py

@@ -14,9 +14,7 @@ Exit codes:
 
 import json
 import os
-import re
 import sys
-from pathlib import Path
 
 # =============================================================================
 # Default constants (overridable via .map/config.yaml → safe_path_prefixes)
@@ -36,6 +34,19 @@ _DEFAULT_DANGEROUS_FILE_PATTERNS = [
     r"tokens?\.(json|ya?ml|toml|txt)$",  # token files, not any file with "token" in path
 ]
 
+_DEFAULT_DANGEROUS_FILE_MARKERS = (
+    ".env",
+    "credential",
+    "private",
+    "key",
+    ".pem",
+    "secret",
+    "id_rsa",
+    "id_ed25519",
+    "password",
+    "token",
+)
+
 # Dangerous bash command patterns
 _DEFAULT_DANGEROUS_COMMANDS = [
     # Block `rm -rf /` (bare root), `rm -rf /etc`, `rm -rf /home/user`, etc.,
@@ -81,14 +92,14 @@ def _load_config_overrides() -> dict:
     Reads safe_path_prefixes from project config to allow customization.
     Falls back to defaults when config is missing or unreadable.
     """
-    project_dir = Path(os.environ.get("CLAUDE_PROJECT_DIR", os.getcwd()))
-    config_path = project_dir / ".map" / "config.yaml"
-    if not config_path.exists():
+    project_dir = os.environ.get("CLAUDE_PROJECT_DIR", os.getcwd())
+    config_path = os.path.join(project_dir, ".map", "config.yaml")
+    if not os.path.exists(config_path):
         return {}
     try:
         import yaml  # type: ignore[import-untyped]
 
-        with open(config_path) as f:
+        with open(config_path, encoding="utf-8") as f:
             data = yaml.safe_load(f)
         return data if isinstance(data, dict) else {}
     except Exception:
@@ -121,6 +132,13 @@ def check_file_safety(path: str) -> tuple[bool, str]:
 
     # Check dangerous patterns
     path_lower = path.lower()
+    if DANGEROUS_FILE_PATTERNS == _DEFAULT_DANGEROUS_FILE_PATTERNS and not any(
+        marker in path_lower for marker in _DEFAULT_DANGEROUS_FILE_MARKERS
+    ):
+        return True, ""
+
+    import re
+
     for pattern in DANGEROUS_FILE_PATTERNS:
         if re.search(pattern, path_lower, re.IGNORECASE):
             return (
@@ -136,6 +154,8 @@ def check_command_safety(command: str) -> tuple[bool, str]:
     if not command:
         return True, ""
 
+    import re
+
     for pattern in DANGEROUS_COMMANDS:
         if re.search(pattern, command, re.IGNORECASE):
             return False, f"Blocked: Dangerous command pattern detected: {pattern}"

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/src/mapify_cli/templates/map/scripts/map_step_runner.py

@@ -5116,6 +5116,68 @@ def _render_review_prompt(
     )
 
 
+def _render_complexity_lens_prompt(
+    review_bundle: str,
+    git_diff: str,
+    minimality_level: str,
+) -> str:
+    """Return the advisory `/map-review` what-to-delete lens prompt."""
+    return "\n\n".join(
+        [
+            "\n".join(
+                [
+                    "<documents>",
+                    "  <document source='.map/<branch>/review-bundle.md' priority='primary'>",
+                    "    <document_content>",
+                    review_bundle,
+                    "    </document_content>",
+                    "  </document>",
+                    "  <document source='git diff' priority='secondary'>",
+                    "    <document_content>",
+                    git_diff,
+                    "    </document_content>",
+                    "  </document>",
+                    "</documents>",
+                ]
+            ),
+            (
+                "<task>\n"
+                "Run the MAP complexity-only what-to-delete lens. "
+                f"Project minimality is {minimality_level}; this lens is disabled when minimality is off.\n"
+                "</task>"
+            ),
+            (
+                "<workflow_policy>\n"
+                "This is advisory-only calibration. Do not gate PROCEED/REVISE/BLOCK on `net: -N`, "
+                "do not create correctness/security/performance findings here, and never feed this output "
+                "into Actor retry context. Normal Monitor/Evaluator review owns blockers.\n"
+                "</workflow_policy>"
+            ),
+            (
+                "<instructions>\n"
+                "Hunt only over-engineering introduced by the current diff. Use exactly these tags:\n"
+                "- delete: dead code, unused flexibility, speculative feature; replacement is nothing.\n"
+                "- stdlib: hand-rolled behavior the standard library ships; name the function.\n"
+                "- native: dependency or code doing what the platform already does; name the feature.\n"
+                "- yagni: abstraction with one implementation, config nobody sets, or layer with one caller.\n"
+                "- shrink: same logic in fewer clear lines; show the shorter form.\n"
+                "Boundaries: complexity only. Correctness bugs, security holes, and performance issues belong "
+                "to normal review, not this lens. A single smoke test or assert-based self-check is the minimum; "
+                "never flag it for deletion. Sample and verify any `map:simplification:` marker claim; the marker "
+                "is evidence, not an exemption.\n"
+                "</instructions>"
+            ),
+            (
+                "<expected_output>\n"
+                "Return plain text only. If cuts exist, write one line per finding exactly as: "
+                "`L<line>: <tag> <what>. <replacement>.` End with exactly: `net: -<N> lines possible.` "
+                "If nothing should be cut, return exactly: `Lean already. Ship.`\n"
+                "</expected_output>"
+            ),
+        ]
+    )
+
+
 def _budget_review_prompt(
     spec: dict[str, str],
     review_bundle: str,
@@ -5153,6 +5215,7 @@ def build_review_prompts(
         else _read_review_bundle_markdown(branch_name)
     )
     git_diff = git_diff_text if git_diff_text is not None else _read_git_diff_for_review()
+    minimality = _load_minimality_level(Path.cwd())
 
     prompts: dict[str, dict[str, object]] = {}
     for role, spec in REVIEW_PROMPT_SPECS.items():
@@ -5167,10 +5230,21 @@ def build_review_prompts(
             "description": spec["description"],
             **prompt_result,
         }
+    if minimality != "off":
+        prompts["complexity_lens"] = {
+            "subagent_type": "evaluator",
+            "description": "Find deletable complexity",
+            "prompt": _render_complexity_lens_prompt(review_bundle, git_diff, minimality),
+            "estimated_tokens": 0,
+            "budget_tokens": budget,
+            "truncated": False,
+            "clipped_sections": [],
+        }
 
     return {
         "status": "success",
         "branch": branch_name,
+        "minimality": minimality,
         "budget_tokens": budget,
         "budget_env": REVIEW_PROMPT_BUDGET_ENV,
         "prompts": prompts,

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/src/mapify_cli/templates/skills/map-review/SKILL.md

@@ -22,7 +22,7 @@ parallel_tool_policy: single_review_fanout
 ```
 
 - Use deeper reasoning for verdicts, risk ranking, section tradeoffs, and contradictory reviewer evidence.
-- Use exactly one parallel reviewer fan-out after bundle preparation: Monitor, Predictor, and Evaluator may run together because they inspect the same review input independently.
+- Use exactly one parallel reviewer fan-out after bundle preparation: Monitor, Predictor, Evaluator, and the optional complexity lens may run together because they inspect the same review input independently.
 - Wait for all reviewer agents before section presentation. Do not parallelize interactive decisions, ordering comparisons that share state, or review-bundle writes.
 
 ## Flags
@@ -287,6 +287,8 @@ REVIEW_PROMPTS_JSON=$(python3 .map/scripts/map_step_runner.py build_review_promp
 MONITOR_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; print(json.load(sys.stdin)["prompts"]["monitor"]["prompt"])')
 PREDICTOR_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; print(json.load(sys.stdin)["prompts"]["predictor"]["prompt"])')
 EVALUATOR_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; print(json.load(sys.stdin)["prompts"]["evaluator"]["prompt"])')
+COMPLEXITY_LENS_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; data=json.load(sys.stdin); print(data.get("prompts",{}).get("complexity_lens",{}).get("prompt", ""))')
+COMPLEXITY_LENS_ENABLED=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; data=json.load(sys.stdin); print("true" if data.get("prompts",{}).get("complexity_lens") else "false")')
 ```
 
 Use the extracted prompt variables as the Task prompts. Keep reviewer task calls below the bundle and prompt-builder commands.
@@ -295,10 +297,13 @@ Use the extracted prompt variables as the Task prompts. Keep reviewer task calls
 Task(subagent_type="monitor", description="Review diff for correctness", prompt=MONITOR_PROMPT)
 Task(subagent_type="predictor", description="Predict integration risk", prompt=PREDICTOR_PROMPT)
 Task(subagent_type="evaluator", description="Score review quality", prompt=EVALUATOR_PROMPT)
+If COMPLEXITY_LENS_ENABLED=true: Task(subagent_type="evaluator", description="Find deletable complexity", prompt=COMPLEXITY_LENS_PROMPT)
 ```
 
 Reviewer prompts reference `review-bundle.json`, `review-bundle.md`, the raw diff as secondary context, and the expected output schema.
 
+When enabled (`minimality != off`), the complexity lens is advisory only. It lists over-engineering as `delete:`, `stdlib:`, `native:`, `yagni:`, or `shrink:` findings, ends with `net: -<N> lines possible.` or `Lean already. Ship.`, samples `map:simplification:` marker claims, and never feeds Actor retries or verdict gates.
+
 ### Step A.2b: Truncated-response gate (MANDATORY — post-fan-out, pre-verification)
 
 After each reviewer returns, validate its output via
@@ -315,6 +320,8 @@ Role → `--agent` kind for the truncation check:
 - predictor reviewer → `--agent predictor`
 - evaluator reviewer → `--agent evaluator`
 
+The optional complexity lens returns plain text, not JSON. Do not run the JSON truncation gate on it; if it is empty or visibly cut off, rerun only that lens prompt once.
+
 ### Step A.3: Verification gate (MANDATORY before any presentation)
 
 For EVERY Monitor / Predictor finding, verify BEFORE listing it as a
@@ -373,6 +380,8 @@ Focus on design boundaries, hidden coupling, state lifecycle, hard/soft constrai
 
 Focus on clarity, duplication, error handling, maintainability, and fit with existing patterns.
 
+If the complexity lens ran, show its raw "what to delete" lines after Code Quality as advisory-only calibration. Do not turn `net: -N` into a REVISE/BLOCK condition.
+
 ### Section: Tests
 
 Focus on changed behavior, failure modes, fixtures, and whether tests prove the contract rather than the implementation.

mapify_cli-3.14.0/src/mapify_cli/templates/skills/map-review/review-reference.md

@@ -0,0 +1,69 @@
+# /map-review Supporting Reference
+
+This file contains lower-frequency review details. Keep `SKILL.md` focused on the active review sequence.
+
+## Section Rubrics
+
+- Architecture: boundaries, lifecycle, coupling, public API behavior, stage consumption.
+- Code Quality: simplicity, naming, duplication, error handling, maintainability.
+- Tests: changed behavior, failure cases, fixtures, coverage of acceptance tags.
+- Performance: hot paths, large artifacts, prompt budgets, avoid speculative micro-optimizations.
+
+## Compare Orderings
+
+When `--compare-orderings` is set, collect one run with `ordering_label='default'`, collect one with `ordering_label='reverse'`, aggregate with `compare-review-runs`, then persist with `record-review-ordering`. Treat verdict drift as review evidence.
+
+## What-To-Delete Lens
+
+When `.map/config.yaml` sets `minimality` to `lite`, `full`, or `ultra`, `build_review_prompts` emits an additional advisory `complexity_lens` prompt. It is deliberately not emitted for `minimality: off` or missing config.
+
+The lens hunts only over-engineering in the current diff and reports one line per finding:
+
+```text
+L<line>: <tag> <what>. <replacement>.
+net: -<N> lines possible.
+```
+
+Allowed tags:
+- `delete:` dead code, unused flexibility, or speculative feature; replacement is nothing.
+- `stdlib:` hand-rolled behavior the standard library already ships; name the function.
+- `native:` dependency or code doing what the platform already does; name the feature.
+- `yagni:` abstraction with one implementation, config nobody sets, or a layer with one caller.
+- `shrink:` same logic in fewer clear lines; show the shorter form.
+
+If nothing should be cut, the entire output is:
+
+```text
+Lean already. Ship.
+```
+
+Boundaries: complexity only. Correctness, security, and performance findings stay in the normal Monitor/Evaluator pass. A single smoke test or assert-based self-check is the minimum and must not be flagged for deletion. The lens samples and verifies `map:simplification:` marker claims; the marker is evidence, not an exemption. `net: -N` is post-hoc and advisory only: do not feed it into Actor retry context, do not use it for PROCEED/REVISE/BLOCK, and do not let it incentivize deleting necessary code.
+
+## Examples
+
+Plain review:
+```text
+/map-review correctness first
+```
+
+Detached review:
+```text
+/map-review --detached
+```
+
+CI review:
+```text
+/map-review --ci
+```
+
+Ordering drift check:
+```text
+/map-review --compare-orderings
+```
+
+## Troubleshooting
+
+- Detached prep unavailable: continue from the in-place review bundle; do not mutate the source branch.
+- Missing bundle: rerun `create_review_bundle` before agents.
+- Prompt clipping: inspect `.map/<branch>/token_budget.json`, then raise `MAP_REVIEW_PROMPT_BUDGET_TOKENS` only when the bundle evidence is actually missing.
+- Monitor invalid: treat as hard stop and record `REVISE` or `BLOCK`.

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/src/mapify_cli/templates_src/hooks/safety-guardrails.py.jinja

@@ -14,9 +14,7 @@ Exit codes:
 
 import json
 import os
-import re
 import sys
-from pathlib import Path
 
 # =============================================================================
 # Default constants (overridable via .map/config.yaml → safe_path_prefixes)
@@ -36,6 +34,19 @@ _DEFAULT_DANGEROUS_FILE_PATTERNS = [
     r"tokens?\.(json|ya?ml|toml|txt)$",  # token files, not any file with "token" in path
 ]
 
+_DEFAULT_DANGEROUS_FILE_MARKERS = (
+    ".env",
+    "credential",
+    "private",
+    "key",
+    ".pem",
+    "secret",
+    "id_rsa",
+    "id_ed25519",
+    "password",
+    "token",
+)
+
 # Dangerous bash command patterns
 _DEFAULT_DANGEROUS_COMMANDS = [
     # Block `rm -rf /` (bare root), `rm -rf /etc`, `rm -rf /home/user`, etc.,
@@ -81,14 +92,14 @@ def _load_config_overrides() -> dict:
     Reads safe_path_prefixes from project config to allow customization.
     Falls back to defaults when config is missing or unreadable.
     """
-    project_dir = Path(os.environ.get("CLAUDE_PROJECT_DIR", os.getcwd()))
-    config_path = project_dir / ".map" / "config.yaml"
-    if not config_path.exists():
+    project_dir = os.environ.get("CLAUDE_PROJECT_DIR", os.getcwd())
+    config_path = os.path.join(project_dir, ".map", "config.yaml")
+    if not os.path.exists(config_path):
         return {}
     try:
         import yaml  # type: ignore[import-untyped]
 
-        with open(config_path) as f:
+        with open(config_path, encoding="utf-8") as f:
             data = yaml.safe_load(f)
         return data if isinstance(data, dict) else {}
     except Exception:
@@ -121,6 +132,13 @@ def check_file_safety(path: str) -> tuple[bool, str]:
 
     # Check dangerous patterns
     path_lower = path.lower()
+    if DANGEROUS_FILE_PATTERNS == _DEFAULT_DANGEROUS_FILE_PATTERNS and not any(
+        marker in path_lower for marker in _DEFAULT_DANGEROUS_FILE_MARKERS
+    ):
+        return True, ""
+
+    import re
+
     for pattern in DANGEROUS_FILE_PATTERNS:
         if re.search(pattern, path_lower, re.IGNORECASE):
             return (
@@ -136,6 +154,8 @@ def check_command_safety(command: str) -> tuple[bool, str]:
     if not command:
         return True, ""
 
+    import re
+
     for pattern in DANGEROUS_COMMANDS:
         if re.search(pattern, command, re.IGNORECASE):
             return False, f"Blocked: Dangerous command pattern detected: {pattern}"

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/src/mapify_cli/templates_src/map/scripts/map_step_runner.py.jinja

@@ -5116,6 +5116,68 @@ def _render_review_prompt(
     )
 
 
+def _render_complexity_lens_prompt(
+    review_bundle: str,
+    git_diff: str,
+    minimality_level: str,
+) -> str:
+    """Return the advisory `/map-review` what-to-delete lens prompt."""
+    return "\n\n".join(
+        [
+            "\n".join(
+                [
+                    "<documents>",
+                    "  <document source='.map/<branch>/review-bundle.md' priority='primary'>",
+                    "    <document_content>",
+                    review_bundle,
+                    "    </document_content>",
+                    "  </document>",
+                    "  <document source='git diff' priority='secondary'>",
+                    "    <document_content>",
+                    git_diff,
+                    "    </document_content>",
+                    "  </document>",
+                    "</documents>",
+                ]
+            ),
+            (
+                "<task>\n"
+                "Run the MAP complexity-only what-to-delete lens. "
+                f"Project minimality is {minimality_level}; this lens is disabled when minimality is off.\n"
+                "</task>"
+            ),
+            (
+                "<workflow_policy>\n"
+                "This is advisory-only calibration. Do not gate PROCEED/REVISE/BLOCK on `net: -N`, "
+                "do not create correctness/security/performance findings here, and never feed this output "
+                "into Actor retry context. Normal Monitor/Evaluator review owns blockers.\n"
+                "</workflow_policy>"
+            ),
+            (
+                "<instructions>\n"
+                "Hunt only over-engineering introduced by the current diff. Use exactly these tags:\n"
+                "- delete: dead code, unused flexibility, speculative feature; replacement is nothing.\n"
+                "- stdlib: hand-rolled behavior the standard library ships; name the function.\n"
+                "- native: dependency or code doing what the platform already does; name the feature.\n"
+                "- yagni: abstraction with one implementation, config nobody sets, or layer with one caller.\n"
+                "- shrink: same logic in fewer clear lines; show the shorter form.\n"
+                "Boundaries: complexity only. Correctness bugs, security holes, and performance issues belong "
+                "to normal review, not this lens. A single smoke test or assert-based self-check is the minimum; "
+                "never flag it for deletion. Sample and verify any `map:simplification:` marker claim; the marker "
+                "is evidence, not an exemption.\n"
+                "</instructions>"
+            ),
+            (
+                "<expected_output>\n"
+                "Return plain text only. If cuts exist, write one line per finding exactly as: "
+                "`L<line>: <tag> <what>. <replacement>.` End with exactly: `net: -<N> lines possible.` "
+                "If nothing should be cut, return exactly: `Lean already. Ship.`\n"
+                "</expected_output>"
+            ),
+        ]
+    )
+
+
 def _budget_review_prompt(
     spec: dict[str, str],
     review_bundle: str,
@@ -5153,6 +5215,7 @@ def build_review_prompts(
         else _read_review_bundle_markdown(branch_name)
     )
     git_diff = git_diff_text if git_diff_text is not None else _read_git_diff_for_review()
+    minimality = _load_minimality_level(Path.cwd())
 
     prompts: dict[str, dict[str, object]] = {}
     for role, spec in REVIEW_PROMPT_SPECS.items():
@@ -5167,10 +5230,21 @@ def build_review_prompts(
             "description": spec["description"],
             **prompt_result,
         }
+    if minimality != "off":
+        prompts["complexity_lens"] = {
+            "subagent_type": "evaluator",
+            "description": "Find deletable complexity",
+            "prompt": _render_complexity_lens_prompt(review_bundle, git_diff, minimality),
+            "estimated_tokens": 0,
+            "budget_tokens": budget,
+            "truncated": False,
+            "clipped_sections": [],
+        }
 
     return {
         "status": "success",
         "branch": branch_name,
+        "minimality": minimality,
         "budget_tokens": budget,
         "budget_env": REVIEW_PROMPT_BUDGET_ENV,
         "prompts": prompts,

{mapify_cli-3.13.1 → mapify_cli-3.14.0}/src/mapify_cli/templates_src/skills/map-review/SKILL.md.jinja

@@ -22,7 +22,7 @@ parallel_tool_policy: single_review_fanout
 ```
 
 - Use deeper reasoning for verdicts, risk ranking, section tradeoffs, and contradictory reviewer evidence.
-- Use exactly one parallel reviewer fan-out after bundle preparation: Monitor, Predictor, and Evaluator may run together because they inspect the same review input independently.
+- Use exactly one parallel reviewer fan-out after bundle preparation: Monitor, Predictor, Evaluator, and the optional complexity lens may run together because they inspect the same review input independently.
 - Wait for all reviewer agents before section presentation. Do not parallelize interactive decisions, ordering comparisons that share state, or review-bundle writes.
 
 ## Flags
@@ -287,6 +287,8 @@ REVIEW_PROMPTS_JSON=$(python3 .map/scripts/map_step_runner.py build_review_promp
 MONITOR_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; print(json.load(sys.stdin)["prompts"]["monitor"]["prompt"])')
 PREDICTOR_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; print(json.load(sys.stdin)["prompts"]["predictor"]["prompt"])')
 EVALUATOR_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; print(json.load(sys.stdin)["prompts"]["evaluator"]["prompt"])')
+COMPLEXITY_LENS_PROMPT=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; data=json.load(sys.stdin); print(data.get("prompts",{}).get("complexity_lens",{}).get("prompt", ""))')
+COMPLEXITY_LENS_ENABLED=$(printf '%s' "$REVIEW_PROMPTS_JSON" | python3 -c 'import json,sys; data=json.load(sys.stdin); print("true" if data.get("prompts",{}).get("complexity_lens") else "false")')
 ```
 
 Use the extracted prompt variables as the Task prompts. Keep reviewer task calls below the bundle and prompt-builder commands.
@@ -295,10 +297,13 @@ Use the extracted prompt variables as the Task prompts. Keep reviewer task calls
 Task(subagent_type="monitor", description="Review diff for correctness", prompt=MONITOR_PROMPT)
 Task(subagent_type="predictor", description="Predict integration risk", prompt=PREDICTOR_PROMPT)
 Task(subagent_type="evaluator", description="Score review quality", prompt=EVALUATOR_PROMPT)
+If COMPLEXITY_LENS_ENABLED=true: Task(subagent_type="evaluator", description="Find deletable complexity", prompt=COMPLEXITY_LENS_PROMPT)
 ```
 
 Reviewer prompts reference `review-bundle.json`, `review-bundle.md`, the raw diff as secondary context, and the expected output schema.
 
+When enabled (`minimality != off`), the complexity lens is advisory only. It lists over-engineering as `delete:`, `stdlib:`, `native:`, `yagni:`, or `shrink:` findings, ends with `net: -<N> lines possible.` or `Lean already. Ship.`, samples `map:simplification:` marker claims, and never feeds Actor retries or verdict gates.
+
 ### Step A.2b: Truncated-response gate (MANDATORY — post-fan-out, pre-verification)
 
 After each reviewer returns, validate its output via
@@ -315,6 +320,8 @@ Role → `--agent` kind for the truncation check:
 - predictor reviewer → `--agent predictor`
 - evaluator reviewer → `--agent evaluator`
 
+The optional complexity lens returns plain text, not JSON. Do not run the JSON truncation gate on it; if it is empty or visibly cut off, rerun only that lens prompt once.
+
 ### Step A.3: Verification gate (MANDATORY before any presentation)
 
 For EVERY Monitor / Predictor finding, verify BEFORE listing it as a
@@ -373,6 +380,8 @@ Focus on design boundaries, hidden coupling, state lifecycle, hard/soft constrai
 
 Focus on clarity, duplication, error handling, maintainability, and fit with existing patterns.
 
+If the complexity lens ran, show its raw "what to delete" lines after Code Quality as advisory-only calibration. Do not turn `net: -N` into a REVISE/BLOCK condition.
+
 ### Section: Tests
 
 Focus on changed behavior, failure modes, fixtures, and whether tests prove the contract rather than the implementation.

mapify_cli-3.14.0/src/mapify_cli/templates_src/skills/map-review/review-reference.md.jinja

@@ -0,0 +1,69 @@
+# /map-review Supporting Reference
+
+This file contains lower-frequency review details. Keep `SKILL.md` focused on the active review sequence.
+
+## Section Rubrics
+
+- Architecture: boundaries, lifecycle, coupling, public API behavior, stage consumption.
+- Code Quality: simplicity, naming, duplication, error handling, maintainability.
+- Tests: changed behavior, failure cases, fixtures, coverage of acceptance tags.
+- Performance: hot paths, large artifacts, prompt budgets, avoid speculative micro-optimizations.
+
+## Compare Orderings
+
+When `--compare-orderings` is set, collect one run with `ordering_label='default'`, collect one with `ordering_label='reverse'`, aggregate with `compare-review-runs`, then persist with `record-review-ordering`. Treat verdict drift as review evidence.
+
+## What-To-Delete Lens
+
+When `.map/config.yaml` sets `minimality` to `lite`, `full`, or `ultra`, `build_review_prompts` emits an additional advisory `complexity_lens` prompt. It is deliberately not emitted for `minimality: off` or missing config.
+
+The lens hunts only over-engineering in the current diff and reports one line per finding:
+
+```text
+L<line>: <tag> <what>. <replacement>.
+net: -<N> lines possible.
+```
+
+Allowed tags:
+- `delete:` dead code, unused flexibility, or speculative feature; replacement is nothing.
+- `stdlib:` hand-rolled behavior the standard library already ships; name the function.
+- `native:` dependency or code doing what the platform already does; name the feature.
+- `yagni:` abstraction with one implementation, config nobody sets, or a layer with one caller.
+- `shrink:` same logic in fewer clear lines; show the shorter form.
+
+If nothing should be cut, the entire output is:
+
+```text
+Lean already. Ship.
+```
+
+Boundaries: complexity only. Correctness, security, and performance findings stay in the normal Monitor/Evaluator pass. A single smoke test or assert-based self-check is the minimum and must not be flagged for deletion. The lens samples and verifies `map:simplification:` marker claims; the marker is evidence, not an exemption. `net: -N` is post-hoc and advisory only: do not feed it into Actor retry context, do not use it for PROCEED/REVISE/BLOCK, and do not let it incentivize deleting necessary code.
+
+## Examples
+
+Plain review:
+```text
+/map-review correctness first
+```
+
+Detached review:
+```text
+/map-review --detached
+```
+
+CI review:
+```text
+/map-review --ci
+```
+
+Ordering drift check:
+```text
+/map-review --compare-orderings
+```
+
+## Troubleshooting
+
+- Detached prep unavailable: continue from the in-place review bundle; do not mutate the source branch.
+- Missing bundle: rerun `create_review_bundle` before agents.
+- Prompt clipping: inspect `.map/<branch>/token_budget.json`, then raise `MAP_REVIEW_PROMPT_BUDGET_TOKENS` only when the bundle evidence is actually missing.
+- Monitor invalid: treat as hard stop and record `REVISE` or `BLOCK`.