specdacular 0.2.2 → 0.2.5

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -1,8 +1,8 @@
1
1
  # Specdacular
2
2
 
3
- **Feature planning for existing codebases.**
3
+ **AI-optimized codebase documentation and feature planning for Claude.**
4
4
 
5
- A Claude Code extension that analyzes your codebase and generates structured documentation. Understand any codebase before planning new features.
5
+ A Claude Code extension that helps you understand codebases and plan features specific enough that an agent can implement without asking questions.
6
6
 
7
7
  ```bash
8
8
  npx specdacular
@@ -12,19 +12,34 @@ npx specdacular
12
12
 
13
13
  ## What It Does
14
14
 
15
- Specdacular spawns 4 parallel AI agents to analyze your codebase and generate 7 structured documents:
15
+ ### 1. Map Your Codebase
16
+
17
+ Spawns 4 parallel agents to analyze your codebase and generate AI-optimized documentation:
16
18
 
17
19
  | Document | What It Answers |
18
20
  |----------|-----------------|
19
- | `STACK.md` | What languages, frameworks, dependencies? |
20
- | `ARCHITECTURE.md` | What patterns? How does data flow? |
21
- | `STRUCTURE.md` | Where do files go? How is it organized? |
22
- | `CONVENTIONS.md` | How should code be written? Naming? Style? |
23
- | `TESTING.md` | How are tests structured? What to mock? |
24
- | `INTEGRATIONS.md` | What external services (DB, APIs, auth)? |
25
- | `CONCERNS.md` | What's broken? Tech debt? Risks? |
21
+ | `MAP.md` | Where is X? What functions exist? |
22
+ | `PATTERNS.md` | How do I write code that fits here? |
23
+ | `STRUCTURE.md` | Where do I put new code? |
24
+ | `CONCERNS.md` | What will bite me? Gotchas? Tech debt? |
25
+
26
+ ### 2. Plan Features
27
+
28
+ A structured flow for planning features with enough detail for agent implementation:
29
+
30
+ ```
31
+ new-feature → discuss-feature → plan-feature →
32
+ (discuss-phase? → research-phase? → execute-plan)* per phase
33
+ ```
26
34
 
27
- Output goes to `.specd/codebase/` in your project.
35
+ **You control the rhythm:**
36
+ - `new-feature` — Initialize and start first discussion
37
+ - `discuss-feature` — Refine understanding (call many times)
38
+ - `research-feature` — Investigate implementation approaches
39
+ - `plan-feature` — Create executable task plans
40
+ - `discuss-phase` — Optional: dive into phase specifics before execution
41
+ - `research-phase` — Optional: research patterns for a specific phase
42
+ - `execute-plan` — Execute plans with progress tracking
28
43
 
29
44
  ---
30
45
 
@@ -35,8 +50,8 @@ npx specdacular
35
50
  ```
36
51
 
37
52
  Choose:
38
- - **Global** (`~/.claude/`) - Available in all projects
39
- - **Local** (`./.claude/`) - This project only
53
+ - **Global** (`~/.claude/`) Available in all projects
54
+ - **Local** (`./.claude/`) This project only
40
55
 
41
56
  ### Verify Installation
42
57
 
@@ -47,36 +62,70 @@ In Claude Code:
47
62
 
48
63
  ---
49
64
 
50
- ## Usage
65
+ ## Commands
66
+
67
+ ### Codebase Documentation
68
+
69
+ | Command | Description |
70
+ |---------|-------------|
71
+ | `/specd:map-codebase` | Analyze codebase with parallel agents |
72
+
73
+ ### Feature Planning
74
+
75
+ | Command | Description |
76
+ |---------|-------------|
77
+ | `/specd:new-feature [name]` | Initialize a feature, start first discussion |
78
+ | `/specd:discuss-feature [name]` | Continue/deepen discussion (iterative) |
79
+ | `/specd:research-feature [name]` | Research implementation with parallel agents |
80
+ | `/specd:plan-feature [name]` | Create executable task plans |
81
+ | `/specd:discuss-phase [feature] [phase]` | Discuss a phase before execution |
82
+ | `/specd:research-phase [feature] [phase]` | Research patterns for a phase |
83
+ | `/specd:execute-plan [feature]` | Execute plans with progress tracking |
84
+
85
+ ### Utilities
86
+
87
+ | Command | Description |
88
+ |---------|-------------|
89
+ | `/specd:help` | Show available commands |
90
+ | `/specd:update` | Update to latest version |
91
+
92
+ ---
93
+
94
+ ## Quick Start
51
95
 
52
- ### Map Your Codebase
96
+ ### Map a Codebase
53
97
 
54
98
  ```
55
99
  /specd:map-codebase
56
100
  ```
57
101
 
58
- This:
59
- 1. Creates `.specd/codebase/` directory
60
- 2. Spawns 4 parallel agents (tech, architecture, quality, concerns)
61
- 3. Each agent explores and writes documents directly
62
- 4. Commits the codebase map
102
+ Creates `.specd/codebase/` with 4 AI-optimized documents.
63
103
 
64
- ### Review Generated Docs
104
+ ### Plan a Feature
65
105
 
66
- ```bash
67
- cat .specd/codebase/STACK.md
68
- cat .specd/codebase/ARCHITECTURE.md
69
- # etc.
106
+ ```
107
+ /specd:new-feature user-dashboard
70
108
  ```
71
109
 
72
- ---
110
+ Creates `.specd/features/user-dashboard/` with:
111
+ - `FEATURE.md` — Technical requirements
112
+ - `CONTEXT.md` — Discussion context (accumulates)
113
+ - `DECISIONS.md` — Decisions with dates and rationale
114
+ - `STATE.md` — Progress tracking
73
115
 
74
- ## Commands
116
+ Then refine and plan:
117
+ ```
118
+ /specd:discuss-feature user-dashboard # Clarify gray areas
119
+ /specd:research-feature user-dashboard # Research implementation
120
+ /specd:plan-feature user-dashboard # Create executable plans
121
+ ```
75
122
 
76
- | Command | Description |
77
- |---------|-------------|
78
- | `/specd:map-codebase` | Analyze codebase and generate documentation |
79
- | `/specd:help` | Show available commands |
123
+ Optionally, before executing each phase:
124
+ ```
125
+ /specd:discuss-phase user-dashboard 1 # Discuss phase 1 specifics
126
+ /specd:research-phase user-dashboard 1 # Research phase 1 patterns
127
+ /specd:execute-plan user-dashboard # Execute with phase context
128
+ ```
80
129
 
81
130
  ---
82
131
 
@@ -84,7 +133,7 @@ cat .specd/codebase/ARCHITECTURE.md
84
133
 
85
134
  ### Parallel Agents
86
135
 
87
- Instead of one agent doing everything, Specdacular spawns 4 specialized agents that run simultaneously:
136
+ Specdacular spawns specialized agents that run simultaneously:
88
137
 
89
138
  ```
90
139
  ┌─────────────────────────────────────────────────────────┐
@@ -94,58 +143,145 @@ Instead of one agent doing everything, Specdacular spawns 4 specialized agents t
94
143
  ┌───────────────┼───────────────┐
95
144
  ▼ ▼ ▼ ▼
96
145
  ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐
97
- Tech │ │ Arch │ │ Quality │ │ Concerns │
146
+ Map │ │ Patterns │ │Structure │ │ Concerns │
98
147
  │ Agent │ │ Agent │ │ Agent │ │ Agent │
99
148
  └──────────┘ └──────────┘ └──────────┘ └──────────┘
100
149
  │ │ │ │
101
150
  ▼ ▼ ▼ ▼
102
- STACK.md ARCH.md CONV.md CONCERNS.md
103
- INTEG.md STRUCT.md TESTING.md
151
+ MAP.md PATTERNS.md STRUCTURE.md CONCERNS.md
104
152
  ```
105
153
 
106
154
  **Benefits:**
107
155
  - Fresh 200k context per agent (no token pollution)
108
156
  - Faster execution (parallel, not sequential)
109
- - Agents write directly to files (minimal context transfer)
157
+ - Agents write directly to files
110
158
 
111
- ---
159
+ ### Feature Flow
112
160
 
113
- ## Updating
161
+ The feature planning flow accumulates context over multiple sessions:
114
162
 
115
- ```bash
116
- npx specdacular@latest
163
+ ```
164
+ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐
165
+ │ new-feature │ ──▶ │ discuss │ ◀─▶ │ research │
166
+ │ │ │ feature │ │ feature │
167
+ └──────────────┘ └──────────────┘ └──────────────┘
168
+
169
+
170
+ ┌──────────────┐
171
+ │ plan-feature │
172
+ └──────────────┘
173
+
174
+
175
+ ┌─────────────────────────────┐
176
+ │ For each phase: │
177
+ │ ┌─────────┐ ┌─────────┐ │
178
+ │ │ discuss │ → │research │ │
179
+ │ │ phase │ │ phase │ │
180
+ │ └─────────┘ └─────────┘ │
181
+ │ │ │ │
182
+ │ └─────┬─────┘ │
183
+ │ ▼ │
184
+ │ ┌────────────┐ │
185
+ │ │execute-plan│ │
186
+ │ └────────────┘ │
187
+ └─────────────────────────────┘
117
188
  ```
118
189
 
119
- ## Uninstalling
190
+ Each session updates:
191
+ - `CONTEXT.md` — Resolved questions accumulate
192
+ - `DECISIONS.md` — Decisions with rationale accumulate
120
193
 
121
- ```bash
122
- npx specdacular --global --uninstall
123
- # or
124
- npx specdacular --local --uninstall
125
- ```
194
+ **Phase-level commands** (optional but powerful):
195
+ - `discuss-phase` Just-in-time clarification for a specific phase
196
+ - `research-phase` — Focused research for phase-specific patterns
197
+
198
+ Plans are prompts for implementing agents with:
199
+ - Specific file paths
200
+ - Code patterns to follow
201
+ - Verification commands
202
+ - Clear completion criteria
126
203
 
127
204
  ---
128
205
 
129
206
  ## Project Structure
130
207
 
131
- After running `/specd:map-codebase`:
208
+ After using Specdacular:
132
209
 
133
210
  ```
134
211
  your-project/
135
212
  ├── .specd/
136
- └── codebase/
137
- ├── STACK.md
138
- ├── ARCHITECTURE.md
139
- ├── STRUCTURE.md
140
- ├── CONVENTIONS.md
141
- ├── TESTING.md
142
- ├── INTEGRATIONS.md
143
- │ └── CONCERNS.md
213
+ ├── codebase/ # From /specd:map-codebase
214
+ ├── MAP.md
215
+ ├── PATTERNS.md
216
+ ├── STRUCTURE.md
217
+ │ └── CONCERNS.md
218
+
219
+ └── features/ # From feature commands
220
+ │ └── user-dashboard/
221
+ │ ├── FEATURE.md # Technical requirements
222
+ │ ├── CONTEXT.md # Feature-level discussion
223
+ │ ├── DECISIONS.md # Decision log (feature + phase)
224
+ │ ├── STATE.md # Progress tracking
225
+ │ ├── RESEARCH.md # Feature-level research
226
+ │ ├── ROADMAP.md # Phase overview
227
+ │ └── plans/ # Executable plans
228
+ │ ├── phase-01/
229
+ │ │ ├── CONTEXT.md # Phase discussion (optional)
230
+ │ │ ├── RESEARCH.md # Phase research (optional)
231
+ │ │ ├── 01-PLAN.md
232
+ │ │ └── 02-PLAN.md
233
+ │ └── phase-02/
234
+ │ ├── CONTEXT.md
235
+ │ ├── RESEARCH.md
236
+ │ └── 01-PLAN.md
144
237
  └── ...
145
238
  ```
146
239
 
147
240
  ---
148
241
 
242
+ ## Philosophy
243
+
244
+ ### Documentation Is For Claude
245
+
246
+ These docs answer questions Claude can't get from reading code:
247
+ - Tribal knowledge
248
+ - Gotchas and pitfalls
249
+ - Patterns and conventions
250
+ - Where things go
251
+
252
+ **Principle:** Don't document what Claude can grep. Document what requires understanding.
253
+
254
+ ### Plans Are Prompts
255
+
256
+ Each `PLAN.md` is literally what you'd send to an implementing agent. It contains everything needed to implement without asking questions.
257
+
258
+ ### Decisions Are Permanent
259
+
260
+ Once recorded in `DECISIONS.md`, decisions aren't re-litigated. Each has date, context, rationale, and implications.
261
+
262
+ ---
263
+
264
+ ## Updating
265
+
266
+ ```bash
267
+ npx specdacular@latest
268
+ ```
269
+
270
+ Or in Claude Code:
271
+ ```
272
+ /specd:update
273
+ ```
274
+
275
+ ## Uninstalling
276
+
277
+ ```bash
278
+ npx specdacular --global --uninstall
279
+ # or
280
+ npx specdacular --local --uninstall
281
+ ```
282
+
283
+ ---
284
+
149
285
  ## License
150
286
 
151
287
  MIT
@@ -0,0 +1,63 @@
1
+ ---
2
+ name: specd:discuss-phase
3
+ description: Discuss a specific phase before execution
4
+ argument-hint: "[feature-name] [phase-number]"
5
+ allowed-tools:
6
+ - Read
7
+ - Write
8
+ - Edit
9
+ - Glob
10
+ - Grep
11
+ - AskUserQuestion
12
+ ---
13
+
14
+ <objective>
15
+ Discuss a specific phase before executing it. Enables just-in-time clarification focused on the phase's scope rather than the entire feature.
16
+
17
+ **Updates:**
18
+ - `.specd/features/{name}/plans/phase-{NN}/CONTEXT.md` — Phase-specific discussion resolutions
19
+ - `.specd/features/{name}/DECISIONS.md` — Accumulates decisions with dates/rationale
20
+
21
+ **Why phase-level discussion?** Instead of researching/discussing the entire feature upfront, dive deeper into each phase's specifics right before executing it. Smaller context, more focused.
22
+ </objective>
23
+
24
+ <execution_context>
25
+ @~/.claude/specdacular/workflows/discuss-phase.md
26
+ </execution_context>
27
+
28
+ <context>
29
+ Arguments: $ARGUMENTS (expects: feature-name phase-number)
30
+
31
+ **Load feature context:**
32
+ @.specd/features/{name}/FEATURE.md
33
+ @.specd/features/{name}/CONTEXT.md
34
+ @.specd/features/{name}/DECISIONS.md
35
+ @.specd/features/{name}/ROADMAP.md
36
+
37
+ **Load phase context:**
38
+ @.specd/features/{name}/plans/phase-{NN}/*.md (plan files)
39
+
40
+ **Load codebase context (if available):**
41
+ @.specd/codebase/PATTERNS.md
42
+ @.specd/codebase/STRUCTURE.md
43
+ @.specd/codebase/MAP.md
44
+ </context>
45
+
46
+ <process>
47
+ 1. **Validate** — Feature exists, phase exists in ROADMAP.md
48
+ 2. **Load Context** — Phase details from ROADMAP.md, plans in phase, feature DECISIONS.md
49
+ 3. **Identify Gray Areas** — Phase-type-specific concerns
50
+ 4. **Probe Gray Areas** — 4 questions per area, then move on
51
+ 5. **Record** — Save to `plans/phase-{NN}/CONTEXT.md`, update DECISIONS.md
52
+ 6. **Commit**
53
+ </process>
54
+
55
+ <success_criteria>
56
+ - [ ] Feature and phase validated
57
+ - [ ] Phase context loaded (plans, goals, files to create/modify)
58
+ - [ ] Phase-type-specific gray areas identified
59
+ - [ ] User-selected areas discussed
60
+ - [ ] Decisions recorded with date, context, rationale
61
+ - [ ] Phase CONTEXT.md created/updated with resolved questions
62
+ - [ ] Committed to git
63
+ </success_criteria>
@@ -0,0 +1,68 @@
1
+ ---
2
+ name: specd:execute-plan
3
+ description: Execute a feature plan with progress tracking
4
+ argument-hint: "[feature-name] [plan-path]"
5
+ allowed-tools:
6
+ - Read
7
+ - Write
8
+ - Edit
9
+ - Bash
10
+ - Glob
11
+ - Grep
12
+ - AskUserQuestion
13
+ ---
14
+
15
+ <objective>
16
+ Execute a plan from a feature, tracking progress and logging deviations.
17
+
18
+ **What it does:**
19
+ 1. Load context — STATE.md, DECISIONS.md, RESEARCH.md, codebase patterns
20
+ 2. Find next plan — First incomplete plan, or accept path as argument
21
+ 3. Execute tasks with:
22
+ - Auto-fix blockers/bugs → log to CHANGELOG.md
23
+ - Ask about architectural changes → wait for user
24
+ - Run verification after each task
25
+ - Stop on verification failure → ask user (retry/skip/stop)
26
+ - Commit after each task
27
+ 4. Update progress — STATE.md with completed tasks
28
+ 5. Suggest next — Next plan to execute
29
+ </objective>
30
+
31
+ <execution_context>
32
+ @~/.claude/specdacular/workflows/execute-plan.md
33
+ </execution_context>
34
+
35
+ <context>
36
+ Feature name: $ARGUMENTS (first argument)
37
+ Plan path: $ARGUMENTS (optional second argument)
38
+
39
+ **Load ALL feature context:**
40
+ @.specd/features/{name}/STATE.md — Progress tracking
41
+ @.specd/features/{name}/DECISIONS.md — Constraints to follow
42
+ @.specd/features/{name}/RESEARCH.md — Implementation notes (if exists)
43
+ @.specd/features/{name}/ROADMAP.md — Phase overview
44
+
45
+ **Load codebase context:**
46
+ @.specd/codebase/PATTERNS.md — Code patterns to follow
47
+ @.specd/codebase/STRUCTURE.md — Where files go
48
+ @.specd/codebase/MAP.md — System overview
49
+ </context>
50
+
51
+ <process>
52
+ 1. **Validate** — Check feature exists with plans
53
+ 2. **Load Context** — Read all feature and codebase docs
54
+ 3. **Find Plan** — Next incomplete or specified plan
55
+ 4. **Execute Tasks** — With verification and deviation handling
56
+ 5. **Complete Plan** — Update STATE.md, suggest next
57
+ </process>
58
+
59
+ <success_criteria>
60
+ - [ ] Feature validated with plans
61
+ - [ ] Context loaded (feature, codebase)
62
+ - [ ] Tasks executed in order
63
+ - [ ] Verification run after each task
64
+ - [ ] Deviations logged to CHANGELOG.md
65
+ - [ ] STATE.md updated with progress
66
+ - [ ] Commits made after each task
67
+ - [ ] Next plan suggested
68
+ </success_criteria>
@@ -30,6 +30,9 @@ Display available specdacular commands and usage guidance.
30
30
  | `/specd:discuss-feature [name]` | Continue/deepen feature discussion (can call many times) |
31
31
  | `/specd:research-feature [name]` | Research implementation with parallel agents |
32
32
  | `/specd:plan-feature [name]` | Create executable task plans for agents |
33
+ | `/specd:discuss-phase [feature] [phase]` | Discuss a phase before execution |
34
+ | `/specd:research-phase [feature] [phase]` | Research patterns for a phase |
35
+ | `/specd:execute-plan [feature] [plan]` | Execute a plan with progress tracking |
33
36
 
34
37
  ### Utilities
35
38
 
@@ -45,7 +48,8 @@ Display available specdacular commands and usage guidance.
45
48
  The feature flow helps you plan features specific enough that an agent can implement without asking questions.
46
49
 
47
50
  ```
48
- new-feature → (discuss ↔ research)* → plan-feature
51
+ new-feature → discuss-feature → plan-feature
52
+ (discuss-phase? → research-phase? → execute-plan)* per phase
49
53
  ```
50
54
 
51
55
  **You control the rhythm:**
@@ -53,6 +57,9 @@ new-feature → (discuss ↔ research)* → plan-feature
53
57
  - `discuss-feature` — Can be called **many times** to refine understanding
54
58
  - `research-feature` — Can be called **many times** to investigate
55
59
  - `plan-feature` — When satisfied, creates executable plans for an agent
60
+ - `discuss-phase` — Optional: dive deeper into phase specifics before execution
61
+ - `research-phase` — Optional: research patterns for a specific phase
62
+ - `execute-plan` — Execute plans with progress tracking and deviation logging
56
63
 
57
64
  ### Quick Start
58
65
 
@@ -63,7 +70,8 @@ new-feature → (discuss ↔ research)* → plan-feature
63
70
  This creates `.specd/features/user-dashboard/` with:
64
71
  - `FEATURE.md` — Technical requirements
65
72
  - `CONTEXT.md` — Discussion context (accumulates)
66
- - `DECISIONS.md` — Decisions with dates and rationale
73
+ - `DECISIONS.md` — User-driven decisions with dates and rationale
74
+ - `CHANGELOG.md` — Auto-captured implementation decisions during execution
67
75
  - `STATE.md` — Progress tracking
68
76
 
69
77
  After initialization, refine with discussion and research, then create plans:
@@ -0,0 +1,63 @@
1
+ ---
2
+ name: specd:research-phase
3
+ description: Research implementation patterns for a phase
4
+ argument-hint: "[feature-name] [phase-number]"
5
+ allowed-tools:
6
+ - Read
7
+ - Write
8
+ - Edit
9
+ - Bash
10
+ - Glob
11
+ - Grep
12
+ - Task
13
+ - AskUserQuestion
14
+ ---
15
+
16
+ <objective>
17
+ Research implementation patterns for a specific phase before executing it. Spawns parallel research agents focused on the phase's scope.
18
+
19
+ **Output:** `.specd/features/{name}/plans/phase-{NN}/RESEARCH.md` with phase-specific guidance.
20
+
21
+ **Why phase-level research?** Instead of researching the entire feature upfront, investigate each phase's specifics right before executing it. Smaller scope, more focused, fresher context.
22
+ </objective>
23
+
24
+ <execution_context>
25
+ @~/.claude/specdacular/workflows/research-phase.md
26
+ </execution_context>
27
+
28
+ <context>
29
+ Arguments: $ARGUMENTS (expects: feature-name phase-number)
30
+
31
+ **Load feature context:**
32
+ @.specd/features/{name}/FEATURE.md
33
+ @.specd/features/{name}/DECISIONS.md
34
+ @.specd/features/{name}/ROADMAP.md
35
+ @.specd/features/{name}/RESEARCH.md (if exists, feature-level research)
36
+
37
+ **Load phase context:**
38
+ @.specd/features/{name}/plans/phase-{NN}/CONTEXT.md (if exists)
39
+ @.specd/features/{name}/plans/phase-{NN}/*.md (plan files)
40
+
41
+ **Load codebase context:**
42
+ @.specd/codebase/PATTERNS.md
43
+ @.specd/codebase/STRUCTURE.md
44
+ @.specd/codebase/MAP.md
45
+ </context>
46
+
47
+ <process>
48
+ 1. **Validate** — Feature and phase exist
49
+ 2. **Load Context** — Phase goals, files to create/modify, previous phase research
50
+ 3. **Spawn Research Agents** — 3 parallel agents for phase-specific research
51
+ 4. **Synthesize** — Combine into `plans/phase-{NN}/RESEARCH.md`
52
+ 5. **Record Decisions** — Technology choices to DECISIONS.md
53
+ 6. **Commit**
54
+ </process>
55
+
56
+ <success_criteria>
57
+ - [ ] Feature and phase validated
58
+ - [ ] Phase context loaded (goals, files, type)
59
+ - [ ] Research agents spawned (codebase, patterns, pitfalls)
60
+ - [ ] Results synthesized into phase RESEARCH.md
61
+ - [ ] Decisions recorded in DECISIONS.md
62
+ - [ ] Committed to git
63
+ </success_criteria>
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "specdacular",
3
- "version": "0.2.2",
3
+ "version": "0.2.5",
4
4
  "description": "Feature planning system for existing codebases. Map, understand, and plan features in large projects.",
5
5
  "bin": {
6
6
  "specdacular": "bin/install.js"
@@ -0,0 +1,34 @@
1
+ # Changelog: {feature-name}
2
+
3
+ Implementation log - auto-captured decisions and deviations during execution.
4
+
5
+ ---
6
+
7
+ ## Phase 1: {Phase Name}
8
+
9
+ ### {YYYY-MM-DD} - Plan 01
10
+
11
+ **{Brief title}**
12
+ - **What:** {What was decided/changed}
13
+ - **Why:** {Reason}
14
+ - **Files:** `{path/to/file}`
15
+
16
+ ---
17
+
18
+ **{Another decision}**
19
+ - **What:** {Description}
20
+ - **Why:** {Reason}
21
+
22
+ ---
23
+
24
+ ## Compacted Summary
25
+
26
+ {Space for human review - summarize key decisions after implementation}
27
+
28
+ ---
29
+
30
+ ## Notes
31
+
32
+ - Entries are auto-captured by executing agent
33
+ - Review after each phase to extract important decisions
34
+ - Move significant decisions to DECISIONS.md if needed
@@ -148,6 +148,33 @@ When this plan is complete:
148
148
 
149
149
  ---
150
150
 
151
+ ## Implementation Log
152
+
153
+ During implementation, capture decisions and deviations to `.specd/features/{feature-name}/CHANGELOG.md`.
154
+
155
+ **When to log:**
156
+ - Choosing a different approach than specified
157
+ - Adding functionality not in the plan
158
+ - Skipping or modifying a task
159
+ - Discovering issues that change the approach
160
+
161
+ **Format:**
162
+ ```markdown
163
+ ### {YYYY-MM-DD} - Plan {NN}
164
+
165
+ **{Brief title}**
166
+ - **What:** {What you decided/changed}
167
+ - **Why:** {Reason for the change}
168
+ - **Files:** `{affected files}`
169
+ ```
170
+
171
+ **Don't log:**
172
+ - Minor implementation details
173
+ - Standard coding patterns
174
+ - Things working as planned
175
+
176
+ ---
177
+
151
178
  ## Notes
152
179
 
153
180
  {Space for the implementing agent to record discoveries, blockers, or decisions made during implementation. Leave empty in template.}
@@ -28,6 +28,19 @@
28
28
 
29
29
  ---
30
30
 
31
+ ## Execution Progress
32
+
33
+ ### Current Plan
34
+ - Plan: none
35
+ - Task: —
36
+ - Started: —
37
+
38
+ ### Completed Plans
39
+ | Plan | Completed | Tasks | Deviations |
40
+ |------|-----------|-------|------------|
41
+
42
+ ---
43
+
31
44
  ## Discussion Sessions
32
45
 
33
46
  | Date | Focus | Outcome |
@@ -65,6 +78,7 @@
65
78
  - `/specd:discuss-feature {feature-name}` — Continue refining understanding
66
79
  - `/specd:research-feature {feature-name}` — Research implementation approach
67
80
  - `/specd:plan-feature {feature-name}` — Create executable plans (when ready)
81
+ - `/specd:execute-plan {feature-name}` — Execute plans with progress tracking
68
82
 
69
83
  ---
70
84
 
@@ -13,5 +13,8 @@
13
13
  "v1_count": 0,
14
14
  "v2_count": 0,
15
15
  "completed": 0
16
+ },
17
+ "execution": {
18
+ "auto_commit": false
16
19
  }
17
20
  }