agent-method 1.5.3 → 1.5.6

package/templates/starter/.context/METHODOLOGY.md

@@ -3,7 +3,7 @@
 Pair with: entry point (CLAUDE.md / .cursorrules / AGENT.md)
 Use when: agent needs operational guidance beyond entry point rules
 
-This file contains methodology operational details that overflow from the entry point. For lite and standard integration profiles, the entry point stays lean — this file provides depth when needed.
+This file contains methodology operational details that overflow from the entry point. For lite and standard integration profiles, the entry point stays lean — this file provides depth when needed. For full profile projects, this file serves as a reference but the entry point contains all rules directly.
 
 ## Workflows
 
@@ -11,8 +11,33 @@ Select based on query type:
 - **General task** — Review, Plan, Implement, Document, Update
 - **Code change** — Review, Plan, Implement, Context-sync, Update
 - **Context refresh** — Scan, Compare, Update-context, Cascade, Record
-- **First session** — Detect, Initialize, Map, Pair, Ready
-- **Discovery (brownfield)** — Inventory, Map-dependencies, Extract-patterns, Assess, Bootstrap
+- **First session (greenfield)** — Ask, Populate, Confirm (see entry point onboarding protocol)
+- **First session (brownfield)** — Scan, Ask, Confirm (see entry point onboarding protocol)
+
+## Checkpoint protocol
+
+The first session protocol in the entry point defines stage-by-stage checkpoints for onboarding. These rules apply to all checkpoints, both during onboarding and normal work:
+
+### When to checkpoint
+
+- **Onboarding**: After each stage's questions are answered, before writing any file
+- **Normal work**: Before executing any plan (checkpoint interaction level)
+- **File writes**: Never write methodology files without showing a preview first
+
+### Checkpoint format
+
+1. Present the draft as a code block the user can review
+2. Wait for explicit approval ("looks good", "yes", "approved")
+3. If the user edits or corrects, incorporate changes and show updated draft
+4. If the user says "skip", leave the file as template placeholder
+
+### Brownfield safeguards
+
+- Stage 1 (Scan) reads only — no writes under any circumstances
+- Existing project files are never modified — methodology files only
+- Entry point customizations are proposed, not applied, until approved
+- Agent presents scan findings as questions, not assertions
+- `.context/REGISTRY.md` is deferred — not created during onboarding (created on first file split to reduce cognitive load on smaller models)
 
 ## Interaction level
 
@@ -56,9 +81,38 @@ When any intelligence layer file exceeds 300 lines (~3,750 tokens):
 3. Components hold archived sections grouped by semantic boundary
 4. Update .context/REGISTRY.md (if it exists) with new file entries
 
-## Audit trail format
+## Reporting layers
+
+Three directories separate reporting concerns from operational files:
+
+### Management/
+
+Management-facing project oversight. Managers read this, not STATE.md or SESSION-LOG.md.
+
+- **DIGEST.md** — Append a row for every high-effort task. Format: date, task, query, outcome, key changes, decisions, effort, time. Low-effort work is NOT recorded.
+- **STATUS.md** — Update after milestones: current phase, active work, recent decisions (last 5), blockers, health indicators.
+
+**What qualifies as high-effort**: Multi-file changes (3+ files), architectural decisions, phase completions, complex multi-step queries, new feature implementations.
+
+**What does NOT qualify**: Status checks, lookups, clarifications, single-file bug fixes, routine maintenance, configuration changes.
+
+### Reviews/
+
+Synthesized views of intelligence layer files. Updated when doing a "project review" query.
+
+- **INDEX.md** — Dashboard with key metrics and links to drill-down files (full tier) or self-contained sections (starter tier).
+- Drill-down files (full tier): roadmap.md, plan.md, project.md, requirements.md, state.md, backlog.md.
+
+### agentWorkflows/
+
+Agent workflow performance analysis. Updated after high-effort sessions.
+
+- **INDEX.md** — Workflow summary, quick stats, links to analysis files (full tier) or self-contained sections (starter tier).
+- Analysis files (full tier): sessions.md (SESSION-LOG synthesis), patterns.md (behavioral patterns), observations.md (case study findings).
+
+### Audit trail format
 
-SUMMARY.md session entries follow this structure:
+SUMMARY.md is the session audit trail. Each entry follows this structure:
 ```
 ### {date} — {brief description}
 **Plan**: {what was planned}

package/templates/starter/.cursorrules

@@ -21,6 +21,10 @@
 | **Methodology guidance** | .context/METHODOLOGY.md | -- |
 | **Phase completion** | SUMMARY.md | SUMMARY.md, STATE.md, ROADMAP.md |
 | **Backlog / ideas** | todos/backlog.md | todos/backlog.md |
+| **Management overview** | Management/DIGEST.md, Management/STATUS.md, STATE.md | Management/STATUS.md, Management/DIGEST.md |
+| **Project review** | Reviews/INDEX.md, relevant source files | Reviews/ (affected review file) |
+| **Workflow analysis** | agentWorkflows/INDEX.md, SESSION-LOG.md | agentWorkflows/ (affected analysis file) |
+| **Docs update** | .context/DOCS-MAP.md, relevant docs/ files | docs/ files, .context/DOCS-MAP.md |
 
 <!-- INSTRUCTION: Replace the {your-domain-type} row with project-specific query types.
      Each should pair .context/BASE.md with one specialist. Add as many rows as needed. -->
@@ -39,18 +43,126 @@ When a file changes, check this table and update dependent files in the same res
 | Intelligence layer file exceeds 300 lines | Restructure into index + components subdirectory (keep active content, archive completed sections) |
 | New domain area | .context/BASE.md, consider new .context/ specialist, this file (if new scoping row) |
 | Lifecycle stage change | PROJECT-PROFILE.md (update stage + date), STATE.md (record decision) |
-| Session close or high-effort task completion | SESSION-LOG.md (append metrics entry — effort, ambiguity, context level, tokens, time, workflow, features, cascades, friction, findings) |
+| Session close or high-effort task completion | SESSION-LOG.md (append metrics entry — effort, ambiguity, context level, tokens, time, user response, refinement delta, workflow, features, cascades, friction, findings). For high-effort tasks also: Management/DIGEST.md (append row — date, task, query, outcome, key changes, decisions, effort, time), Management/STATUS.md (if milestone) |
+| Significant milestone or phase completion | Management/STATUS.md (update current state, recent decisions, health indicators) |
+| New project component or module | .context/DOCS-MAP.md (add component mapping, check scaffolding rules for new docs/ proposals) |
+| docs/ file created or deleted | .context/DOCS-MAP.md (update inventory), docs/index.md (update navigation — create if it doesn't exist yet) |
 
 <!-- INSTRUCTION: Add project-specific cascade rules below the universal ones above. -->
 
+## Docs maintenance
+
+<!-- INSTRUCTION: This table tells the agent when to update human-facing docs.
+     For full component-to-docs mapping, see .context/DOCS-MAP.md. -->
+
+| Trigger | Update these docs/ files |
+|---------|------------------------|
+| Phase completed | docs/index.md (progress section — create from DOCS-MAP.md if it doesn't exist yet) |
+| Architecture decision | docs/index.md (architecture section), .context/DOCS-MAP.md (if new component) |
+| New project component | .context/DOCS-MAP.md (add mapping), docs/ (per DOCS-MAP.md scaffolding rules) |
+| {structural-change} | {specific docs/ files per DOCS-MAP.md component mapping} |
+
 ## Workflow
 
 Select based on query type:
 - **General task** — Review, Plan, Implement, Document, Update
 - **Code change** — Review, Plan, Implement, Context-sync, Update
 - **Context refresh** — Scan, Compare, Update-context, Cascade, Record
-- **First session** — Detect, Initialize, Map, Pair, Ready
-- **Discovery (brownfield)** — Inventory, Map-dependencies, Extract-patterns, Assess, Bootstrap
+- **First session (greenfield)** — Ask, Populate, Confirm (see onboarding protocol below)
+- **First session (brownfield)** — Scan, Ask, Confirm (see onboarding protocol below)
+
+## First session protocol
+
+<!-- AGENT: Follow this protocol on the FIRST session when methodology files contain only template placeholders. -->
+
+When PROJECT.md still contains `{placeholder}` text, this is a first session. Guide the user through onboarding — ask questions, show what you'll write, get confirmation before writing each file.
+
+### Greenfield onboarding (no existing codebase)
+
+**Stage 1 — Project vision** → populates PROJECT.md
+Ask the user:
+1. What is your project about? (1-2 sentences)
+2. Who is it for?
+3. What problem does it solve?
+4. What technologies, languages, or frameworks will you use?
+5. What are 2-3 guiding principles for this project?
+
+CHECKPOINT: Draft PROJECT.md from the answers. Show the user the draft. Wait for approval before writing.
+
+**Stage 2 — Architecture** → populates .context/BASE.md, PROJECT-PROFILE.md, .context/DOCS-MAP.md
+Ask the user:
+1. What are the main components or modules?
+2. How do they connect? (data flow, API calls, event-driven, etc.)
+3. What is the directory structure? (or what will it be?)
+4. What are the key dependencies?
+
+CHECKPOINT: Draft .context/BASE.md, PROJECT-PROFILE.md, and .context/DOCS-MAP.md (initial component-to-docs mapping based on architecture). Show the user. Wait for approval.
+
+**Stage 3 — Goals** → populates ROADMAP.md, PLAN.md
+Ask the user:
+1. What are your immediate goals? (next 1-3 sessions)
+2. What does "done" look like for this project?
+3. Can you break this into 2-4 phases?
+
+CHECKPOINT: Draft ROADMAP.md and PLAN.md (first task). Show the user. Wait for approval.
+
+**Stage 4 — Starting state** → populates STATE.md
+Ask the user:
+1. Any decisions already made? (tech choices, architecture, constraints)
+2. Any known blockers or risks?
+3. Any open questions you want tracked?
+
+CHECKPOINT: Draft STATE.md. Show the user. Wait for approval. Then begin normal work.
+
+### Brownfield onboarding (existing codebase)
+
+**Stage 1 — Scan** → reads existing files, writes nothing
+Read: package.json, README, directory structure, key config files, entry points.
+Present findings: "Here's what I found in your project: [summary]. Is this accurate?"
+
+CHECKPOINT: Wait for user to confirm or correct your understanding.
+
+**Stage 2 — Clarify** → populates PROJECT.md, PROJECT-PROFILE.md
+Ask targeted questions based on what the scan missed:
+1. What's the intent behind [pattern you found]?
+2. Are there conventions not captured in config? (naming, structure, etc.)
+3. What's the current state — active development, maintenance, exploration?
+
+CHECKPOINT: Draft PROJECT.md from scan + answers. Show the user. Wait for approval.
+
+**Stage 3 — Map** → populates .context/BASE.md, .context/DOCS-MAP.md
+Build codebase map from scan. Ask:
+1. Are there areas of the codebase I should understand deeply vs. ignore?
+2. Any architectural decisions I should know about that aren't documented?
+3. What are the most common types of changes you make?
+
+CHECKPOINT: Draft .context/BASE.md and .context/DOCS-MAP.md (component-to-docs mapping from scan results). Show the user. Wait for approval.
+
+**Stage 4 — Organize** → proposes entry point customization
+Based on understanding, propose:
+1. Scoping rules — which query types match this project?
+2. Specialist context files — which domains need their own .context/ file?
+3. Cascade rules — what project-specific dependencies exist?
+
+CHECKPOINT: Show proposed entry point customizations. Wait for approval before modifying this file.
+
+**Stage 5 — Starting state** → populates STATE.md, ROADMAP.md
+Propose initial state based on everything learned:
+1. Decisions already implicit in the codebase (document them)
+2. Current position (where is the project in its lifecycle?)
+3. Immediate goals
+
+CHECKPOINT: Draft STATE.md and ROADMAP.md. Show the user. Wait for approval. Then begin normal work.
+
+### Checkpoint rules
+
+- NEVER write to a methodology file without showing the user what you will write first
+- At each CHECKPOINT, present the draft as a code block the user can review
+- If the user says "looks good" or similar, write the file
+- If the user edits or corrects, incorporate changes and show the updated draft
+- If the user says "skip", leave the file as template placeholder
+- After all stages complete, summarize what was populated and begin normal work
+- For brownfield: `.context/REGISTRY.md` is deferred — do not create it during onboarding. It will be created automatically when the first file split occurs (reduces first-session cognitive load)
 
 ## Interaction level
 
@@ -70,21 +182,28 @@ tier: standard
 
 method_version: 1.5
 <!-- Tracks which methodology version generated this entry point -->
-<!-- Use `npx wwa status` to compare against latest -->
+<!-- Use `wwa status` to compare against latest -->
 
 ## CLI tools (optional)
 
-Available via `npx wwa` (zero-install) or `pip install wwa-tools`:
+Install: `npm install -g agent-method` (then use `wwa`), or `pip install wwa-tools`.
 
 | When you want to... | Run |
 |---------------------|-----|
-| Validate this entry point | `npx wwa check` |
-| See what type of project this is | `npx wwa scan` |
-| Test how a query routes | `npx wwa route "your question"` |
-| Extract a refinement report | `npx wwa refine` |
-| Check methodology version | `npx wwa status` |
-| Update methodology files | `npx wwa upgrade` |
-| See what an entry point should contain | `npx wwa init code` / `context` / `data` / `mix` |
+| Validate this entry point | `wwa check` |
+| See what type of project this is | `wwa scan` |
+| Test how a query routes | `wwa route "your question"` |
+| See current plan status | `wwa plan` |
+| Get implementation guidance | `wwa implement` |
+| View project review dashboard | `wwa review` |
+| Show management digest | `wwa digest` |
+| Session close + project snapshot | `wwa close` |
+| Extract a refinement report | `wwa refine` |
+| Extract a case study (with project tokens) | `wwa casestudy` |
+| Check docs coverage | `wwa docs` |
+| Check methodology version | `wwa status` |
+| Update methodology files | `wwa upgrade` |
+| Set up a new project | `wwa init code` / `context` / `data` / `mix` |
 
 <!-- INSTRUCTION: The agent can suggest these commands when the user asks about validation,
      project setup, or methodology updates. All commands auto-detect project type and find
@@ -99,7 +218,9 @@ Available via `npx wwa` (zero-install) or `pip install wwa-tools`:
 - Surface uncertainty as open questions in STATE.md — never guess silently
 - Keep intelligence layer files under 300 lines — split into index + components subdirectory when exceeded
 - Propose plans and wait for approval — the human controls direction
-- At session close or after any high-effort task, append a metrics entry to SESSION-LOG.md — include effort level, question ambiguity, context level, estimated tokens, and time. Never skip, never read previous entries during normal work
+- Document proportionally — low-effort tasks (status checks, lookups, clarifications) need no STATE.md or SESSION-LOG updates; high-effort tasks get full recording
+- SUMMARY.md is the session audit trail (date, plan, outcome, files, decisions, next). Management digest lives in Management/DIGEST.md
+- At session close or after any high-effort task, append a metrics entry to SESSION-LOG.md — include effort level, question ambiguity, context level, estimated tokens, time, user response (accepted/edited/revised/rejected/redirected), and for medium/high effort tasks: revision count, refinement magnitude (none/minor/moderate/major/rework), delta categories, and survival rate. Never skip, never read previous entries during normal work
 
 ## Do not
 
@@ -109,5 +230,6 @@ Available via `npx wwa` (zero-install) or `pip install wwa-tools`:
 - Defer STATE.md decision recording to end of session
 - Skip cascade checks after file changes
 - Load multiple specialist .context/ files for a single query
+- Record routine queries in STATE.md or SESSION-LOG.md — only decisions, blockers, and high-effort work warrant documentation
 
 <!-- INSTRUCTION: Add project-specific "do not" rules as you discover common mistakes -->

package/templates/starter/AGENT.md

@@ -21,6 +21,10 @@
 | **Methodology guidance** | .context/METHODOLOGY.md | -- |
 | **Phase completion** | SUMMARY.md | SUMMARY.md, STATE.md, ROADMAP.md |
 | **Backlog / ideas** | todos/backlog.md | todos/backlog.md |
+| **Management overview** | Management/DIGEST.md, Management/STATUS.md, STATE.md | Management/STATUS.md, Management/DIGEST.md |
+| **Project review** | Reviews/INDEX.md, relevant source files | Reviews/ (affected review file) |
+| **Workflow analysis** | agentWorkflows/INDEX.md, SESSION-LOG.md | agentWorkflows/ (affected analysis file) |
+| **Docs update** | .context/DOCS-MAP.md, relevant docs/ files | docs/ files, .context/DOCS-MAP.md |
 
 <!-- INSTRUCTION: Replace the {your-domain-type} row with project-specific query types.
      Each should pair .context/BASE.md with one specialist. Add as many rows as needed. -->
@@ -39,18 +43,126 @@ When a file changes, check this table and update dependent files in the same res
 | Intelligence layer file exceeds 300 lines | Restructure into index + components subdirectory (keep active content, archive completed sections) |
 | New domain area | .context/BASE.md, consider new .context/ specialist, this file (if new scoping row) |
 | Lifecycle stage change | PROJECT-PROFILE.md (update stage + date), STATE.md (record decision) |
-| Session close or high-effort task completion | SESSION-LOG.md (append metrics entry — effort, ambiguity, context level, tokens, time, workflow, features, cascades, friction, findings) |
+| Session close or high-effort task completion | SESSION-LOG.md (append metrics entry — effort, ambiguity, context level, tokens, time, user response, refinement delta, workflow, features, cascades, friction, findings). For high-effort tasks also: Management/DIGEST.md (append row — date, task, query, outcome, key changes, decisions, effort, time), Management/STATUS.md (if milestone) |
+| Significant milestone or phase completion | Management/STATUS.md (update current state, recent decisions, health indicators) |
+| New project component or module | .context/DOCS-MAP.md (add component mapping, check scaffolding rules for new docs/ proposals) |
+| docs/ file created or deleted | .context/DOCS-MAP.md (update inventory), docs/index.md (update navigation — create if it doesn't exist yet) |
 
 <!-- INSTRUCTION: Add project-specific cascade rules below the universal ones above. -->
 
+## Docs maintenance
+
+<!-- INSTRUCTION: This table tells the agent when to update human-facing docs.
+     For full component-to-docs mapping, see .context/DOCS-MAP.md. -->
+
+| Trigger | Update these docs/ files |
+|---------|------------------------|
+| Phase completed | docs/index.md (progress section — create from DOCS-MAP.md if it doesn't exist yet) |
+| Architecture decision | docs/index.md (architecture section), .context/DOCS-MAP.md (if new component) |
+| New project component | .context/DOCS-MAP.md (add mapping), docs/ (per DOCS-MAP.md scaffolding rules) |
+| {structural-change} | {specific docs/ files per DOCS-MAP.md component mapping} |
+
 ## Workflow
 
 Select based on query type:
 - **General task** — Review, Plan, Implement, Document, Update
 - **Code change** — Review, Plan, Implement, Context-sync, Update
 - **Context refresh** — Scan, Compare, Update-context, Cascade, Record
-- **First session** — Detect, Initialize, Map, Pair, Ready
-- **Discovery (brownfield)** — Inventory, Map-dependencies, Extract-patterns, Assess, Bootstrap
+- **First session (greenfield)** — Ask, Populate, Confirm (see onboarding protocol below)
+- **First session (brownfield)** — Scan, Ask, Confirm (see onboarding protocol below)
+
+## First session protocol
+
+<!-- AGENT: Follow this protocol on the FIRST session when methodology files contain only template placeholders. -->
+
+When PROJECT.md still contains `{placeholder}` text, this is a first session. Guide the user through onboarding — ask questions, show what you'll write, get confirmation before writing each file.
+
+### Greenfield onboarding (no existing codebase)
+
+**Stage 1 — Project vision** → populates PROJECT.md
+Ask the user:
+1. What is your project about? (1-2 sentences)
+2. Who is it for?
+3. What problem does it solve?
+4. What technologies, languages, or frameworks will you use?
+5. What are 2-3 guiding principles for this project?
+
+CHECKPOINT: Draft PROJECT.md from the answers. Show the user the draft. Wait for approval before writing.
+
+**Stage 2 — Architecture** → populates .context/BASE.md, PROJECT-PROFILE.md, .context/DOCS-MAP.md
+Ask the user:
+1. What are the main components or modules?
+2. How do they connect? (data flow, API calls, event-driven, etc.)
+3. What is the directory structure? (or what will it be?)
+4. What are the key dependencies?
+
+CHECKPOINT: Draft .context/BASE.md, PROJECT-PROFILE.md, and .context/DOCS-MAP.md (initial component-to-docs mapping based on architecture). Show the user. Wait for approval.
+
+**Stage 3 — Goals** → populates ROADMAP.md, PLAN.md
+Ask the user:
+1. What are your immediate goals? (next 1-3 sessions)
+2. What does "done" look like for this project?
+3. Can you break this into 2-4 phases?
+
+CHECKPOINT: Draft ROADMAP.md and PLAN.md (first task). Show the user. Wait for approval.
+
+**Stage 4 — Starting state** → populates STATE.md
+Ask the user:
+1. Any decisions already made? (tech choices, architecture, constraints)
+2. Any known blockers or risks?
+3. Any open questions you want tracked?
+
+CHECKPOINT: Draft STATE.md. Show the user. Wait for approval. Then begin normal work.
+
+### Brownfield onboarding (existing codebase)
+
+**Stage 1 — Scan** → reads existing files, writes nothing
+Read: package.json, README, directory structure, key config files, entry points.
+Present findings: "Here's what I found in your project: [summary]. Is this accurate?"
+
+CHECKPOINT: Wait for user to confirm or correct your understanding.
+
+**Stage 2 — Clarify** → populates PROJECT.md, PROJECT-PROFILE.md
+Ask targeted questions based on what the scan missed:
+1. What's the intent behind [pattern you found]?
+2. Are there conventions not captured in config? (naming, structure, etc.)
+3. What's the current state — active development, maintenance, exploration?
+
+CHECKPOINT: Draft PROJECT.md from scan + answers. Show the user. Wait for approval.
+
+**Stage 3 — Map** → populates .context/BASE.md, .context/DOCS-MAP.md
+Build codebase map from scan. Ask:
+1. Are there areas of the codebase I should understand deeply vs. ignore?
+2. Any architectural decisions I should know about that aren't documented?
+3. What are the most common types of changes you make?
+
+CHECKPOINT: Draft .context/BASE.md and .context/DOCS-MAP.md (component-to-docs mapping from scan results). Show the user. Wait for approval.
+
+**Stage 4 — Organize** → proposes entry point customization
+Based on understanding, propose:
+1. Scoping rules — which query types match this project?
+2. Specialist context files — which domains need their own .context/ file?
+3. Cascade rules — what project-specific dependencies exist?
+
+CHECKPOINT: Show proposed entry point customizations. Wait for approval before modifying this file.
+
+**Stage 5 — Starting state** → populates STATE.md, ROADMAP.md
+Propose initial state based on everything learned:
+1. Decisions already implicit in the codebase (document them)
+2. Current position (where is the project in its lifecycle?)
+3. Immediate goals
+
+CHECKPOINT: Draft STATE.md and ROADMAP.md. Show the user. Wait for approval. Then begin normal work.
+
+### Checkpoint rules
+
+- NEVER write to a methodology file without showing the user what you will write first
+- At each CHECKPOINT, present the draft as a code block the user can review
+- If the user says "looks good" or similar, write the file
+- If the user edits or corrects, incorporate changes and show the updated draft
+- If the user says "skip", leave the file as template placeholder
+- After all stages complete, summarize what was populated and begin normal work
+- For brownfield: `.context/REGISTRY.md` is deferred — do not create it during onboarding. It will be created automatically when the first file split occurs (reduces first-session cognitive load)
 
 ## Interaction level
 
@@ -70,21 +182,28 @@ tier: standard
 
 method_version: 1.5
 <!-- Tracks which methodology version generated this entry point -->
-<!-- Use `npx wwa status` to compare against latest -->
+<!-- Use `wwa status` to compare against latest -->
 
 ## CLI tools (optional)
 
-Available via `npx wwa` (zero-install) or `pip install wwa-tools`:
+Install: `npm install -g agent-method` (then use `wwa`), or `pip install wwa-tools`.
 
 | When you want to... | Run |
 |---------------------|-----|
-| Validate this entry point | `npx wwa check` |
-| See what type of project this is | `npx wwa scan` |
-| Test how a query routes | `npx wwa route "your question"` |
-| Extract a refinement report | `npx wwa refine` |
-| Check methodology version | `npx wwa status` |
-| Update methodology files | `npx wwa upgrade` |
-| See what an entry point should contain | `npx wwa init code` / `context` / `data` / `mix` |
+| Validate this entry point | `wwa check` |
+| See what type of project this is | `wwa scan` |
+| Test how a query routes | `wwa route "your question"` |
+| See current plan status | `wwa plan` |
+| Get implementation guidance | `wwa implement` |
+| View project review dashboard | `wwa review` |
+| Show management digest | `wwa digest` |
+| Session close + project snapshot | `wwa close` |
+| Extract a refinement report | `wwa refine` |
+| Extract a case study (with project tokens) | `wwa casestudy` |
+| Check docs coverage | `wwa docs` |
+| Check methodology version | `wwa status` |
+| Update methodology files | `wwa upgrade` |
+| Set up a new project | `wwa init code` / `context` / `data` / `mix` |
 
 <!-- INSTRUCTION: The agent can suggest these commands when the user asks about validation,
      project setup, or methodology updates. All commands auto-detect project type and find
@@ -99,7 +218,9 @@ Available via `npx wwa` (zero-install) or `pip install wwa-tools`:
 - Surface uncertainty as open questions in STATE.md — never guess silently
 - Keep intelligence layer files under 300 lines — split into index + components subdirectory when exceeded
 - Propose plans and wait for approval — the human controls direction
-- At session close or after any high-effort task, append a metrics entry to SESSION-LOG.md — include effort level, question ambiguity, context level, estimated tokens, and time. Never skip, never read previous entries during normal work
+- Document proportionally — low-effort tasks (status checks, lookups, clarifications) need no STATE.md or SESSION-LOG updates; high-effort tasks get full recording
+- SUMMARY.md is the session audit trail (date, plan, outcome, files, decisions, next). Management digest lives in Management/DIGEST.md
+- At session close or after any high-effort task, append a metrics entry to SESSION-LOG.md — include effort level, question ambiguity, context level, estimated tokens, time, user response (accepted/edited/revised/rejected/redirected), and for medium/high effort tasks: revision count, refinement magnitude (none/minor/moderate/major/rework), delta categories, and survival rate. Never skip, never read previous entries during normal work
 
 ## Do not
 
@@ -109,5 +230,6 @@ Available via `npx wwa` (zero-install) or `pip install wwa-tools`:
 - Defer STATE.md decision recording to end of session
 - Skip cascade checks after file changes
 - Load multiple specialist .context/ files for a single query
+- Record routine queries in STATE.md or SESSION-LOG.md — only decisions, blockers, and high-effort work warrant documentation
 
 <!-- INSTRUCTION: Add project-specific "do not" rules as you discover common mistakes -->