generate

Name: generate
Author: anombyte93

582stars57forks

How to install generate?

anombyte93/prd-taskmaster/generate

$npx -y skills add anombyte93/prd-taskmaster --skill generate

Installs into the current project.

Summary

Phase 2 of the prd-taskmaster pipeline: spec generation and task parsing. Loads a template (comprehensive|minimal), fills it with DISCOVER-phase constraints and answers, validates the spec (placeholders_found, grade thresholds), parses the PRD into tasks via task-master, runs Tas

Files · 1

View on GitHub

SKILL.md

1	# Phase 2: Generate
2
3	Declarative phase skill. Invoked by the prd-taskmaster orchestrator when
4	`current_phase` is `GENERATE`. Never called directly by a user.
5
6	The one rule: **generate the spec, validate it catches placeholders, parse it
7	into tasks, expand every task into subtasks. Quality over speed.**
8
9	## Entry gate
10
11	1. Call `mcp__plugin_prd_go__check_gate(phase="GENERATE", evidence={})` for diagnostics.
12
13	`check_gate` is an EXIT gate: it checks `task_count > 0`, `subtask_coverage >= 1.0`,
14	and `validation_grade in (EXCELLENT, GOOD)` — all of which are GENERATE's OWN OUTPUTS,
15	i.e. evidence to advance, not preconditions to enter. On first entry none exist
16	yet, so a `gate_passed: false` here is EXPECTED — the state machine's legal
17	transitions already guarantee only legal entry.
18
19	- First entry (no evidence yet): note the result and continue with the Procedure.
20	- Re-entry: if the gate reports violations, report them and stop — it protects
21	against re-running a completed phase or skipping ahead from DISCOVER.
22	2. Read the DISCOVER output (discovery summary + `CONSTRAINTS CAPTURED` block
23	+ scale classification). If any of these are missing, report and stop — the
24	gate should have caught this, but belt-and-braces.
25
26	## Generate checklist
27
28	Copy into your response before running the procedure:
29
30	```
31	GENERATE CHECKLIST:
32	- [ ] Template loaded (comprehensive\|minimal)
33	- [ ] Spec written with discovery answers (no bare placeholders remaining)
34	- [ ] CONSTRAINT CHECK: every DISCOVER constraint appears in the spec
35	- [ ] SCOPE CHECK: task count matches scale (Solo 8-12, Team 12-20, Enterprise 20-30)
36	- [ ] Validation score: ___ / ___ (grade: ___)
37	- [ ] placeholders_found: ___ (bare placeholders = 0 required)
38	- [ ] Warnings addressed or acknowledged
39	- [ ] Tasks parsed: ___ tasks created
40	- [ ] Complexity analyzed via TaskMaster: Y/N
41	- [ ] All tasks expanded into subtasks: Y/N
42	```
43
44	## Step 1: Choose and load template
45
46	Decide based on discovery depth:
47
48	- Comprehensive: 4+ detailed answers, complex project, Team / Enterprise scale
49	- Minimal: thin answers, user wants speed, Solo scale
50
51	MCP (preferred): `mcp__plugin_prd_go__load_template(type="comprehensive")`
52
53	CLI fallback: `python3 script.py load-template --type comprehensive`
54
55	The template is the canonical shape — do not invent your own. If the template
56	load fails, report and stop. Do not paper over with a home-rolled skeleton.
57
58	## Step 2: Generate spec at `.taskmaster/docs/prd.md`
59
60	Fill the template with discovery answers. AI judgment required:
61
62	- Replace ALL placeholders with actual content pulled from DISCOVER.
63	- Expand with project-specific details — do not leave template prose verbatim.
64	- Add technical depth proportional to what the user provided in discovery.
65	- Generate domain-appropriate sections (pentest = threat model, app = user
66	stories, business = success metrics, learning = assessment criteria).
67	- Document assumptions where discovery was thin — explicitly, not silently.
68
69	### CONSTRAINT CHECK (MANDATORY)
70
71	Verify EVERY constraint from the DISCOVER phase `CONSTRAINTS CAPTURED` block
72	appears in the spec. If "must use Python" was a constraint, the spec MUST
73	reference Python. Missing constraints = spec bug.
74
75	Emit the check explicitly:
76
77	```
78	CONSTRAINT CHECK:
79	- Tech stack (Python): FOUND in spec section "Technical Stack"
80	- Timeline (MVP in 2 weeks): FOUND in spec section "Milestones"
81	- ...
82	```
83
84	Every constraint must be marked FOUND. If any are MISSING, loop back and
85	fix the spec before proceeding.
86
87	### SCOPE CHECK (MANDATORY)
88
89	Use the scale classification from DISCOVER to set task count range:
90
91	\| Scale \| Task Count \| Subtask Depth \|
92	\|------------\|-----------\|------------------\|
93	\| Solo \| 8–12 \| 2–3 subtasks each \|
94	\| Team \| 12–20 \| 3–5 subtasks each \|
95	\| Enterprise \| 20–30 \| 5–8 subtasks each \|
96
97	If DISCOVER classified the project as Team but the spec implies 30 tasks,
98	that's a scope bug — narrow the spec or re-classify explicitly.
99
100	### Domain-neutral vocabulary
101
102	When the domain is unclear, default to neutral terms:
103
104	\| Software term \| Neutral equivalent \| When to use neutral \|
105	\|---------------\|-------------------\|---------------------\|
106	\| tests \| verification criteria \| pentest, business, learning \|
107	\| code

People who liked this also viewed

6 picks

Type

generate

582stars57forks

View on GitHub

How to install generate?

anombyte93/prd-taskmaster/generate

$npx -y skills add anombyte93/prd-taskmaster --skill generate

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Phase 2: Generate
2
3	Declarative phase skill. Invoked by the prd-taskmaster orchestrator when
4	`current_phase` is `GENERATE`. Never called directly by a user.
5
6	The one rule: **generate the spec, validate it catches placeholders, parse it
7	into tasks, expand every task into subtasks. Quality over speed.**
8
9	## Entry gate
10
11	1. Call `mcp__plugin_prd_go__check_gate(phase="GENERATE", evidence={})` for diagnostics.
12
13	`check_gate` is an EXIT gate: it checks `task_count > 0`, `subtask_coverage >= 1.0`,
14	and `validation_grade in (EXCELLENT, GOOD)` — all of which are GENERATE's OWN OUTPUTS,
15	i.e. evidence to advance, not preconditions to enter. On first entry none exist
16	yet, so a `gate_passed: false` here is EXPECTED — the state machine's legal
17	transitions already guarantee only legal entry.
18
19	- First entry (no evidence yet): note the result and continue with the Procedure.
20	- Re-entry: if the gate reports violations, report them and stop — it protects
21	against re-running a completed phase or skipping ahead from DISCOVER.
22	2. Read the DISCOVER output (discovery summary + `CONSTRAINTS CAPTURED` block
23	+ scale classification). If any of these are missing, report and stop — the
24	gate should have caught this, but belt-and-braces.
25
26	## Generate checklist
27
28	Copy into your response before running the procedure:
29
30	```
31	GENERATE CHECKLIST:
32	- [ ] Template loaded (comprehensive\|minimal)
33	- [ ] Spec written with discovery answers (no bare placeholders remaining)
34	- [ ] CONSTRAINT CHECK: every DISCOVER constraint appears in the spec
35	- [ ] SCOPE CHECK: task count matches scale (Solo 8-12, Team 12-20, Enterprise 20-30)
36	- [ ] Validation score: ___ / ___ (grade: ___)
37	- [ ] placeholders_found: ___ (bare placeholders = 0 required)
38	- [ ] Warnings addressed or acknowledged
39	- [ ] Tasks parsed: ___ tasks created
40	- [ ] Complexity analyzed via TaskMaster: Y/N
41	- [ ] All tasks expanded into subtasks: Y/N
42	```
43
44	## Step 1: Choose and load template
45
46	Decide based on discovery depth:
47
48	- Comprehensive: 4+ detailed answers, complex project, Team / Enterprise scale
49	- Minimal: thin answers, user wants speed, Solo scale
50
51	MCP (preferred): `mcp__plugin_prd_go__load_template(type="comprehensive")`
52
53	CLI fallback: `python3 script.py load-template --type comprehensive`
54
55	The template is the canonical shape — do not invent your own. If the template
56	load fails, report and stop. Do not paper over with a home-rolled skeleton.
57
58	## Step 2: Generate spec at `.taskmaster/docs/prd.md`
59
60	Fill the template with discovery answers. AI judgment required:
61
62	- Replace ALL placeholders with actual content pulled from DISCOVER.
63	- Expand with project-specific details — do not leave template prose verbatim.
64	- Add technical depth proportional to what the user provided in discovery.
65	- Generate domain-appropriate sections (pentest = threat model, app = user
66	stories, business = success metrics, learning = assessment criteria).
67	- Document assumptions where discovery was thin — explicitly, not silently.
68
69	### CONSTRAINT CHECK (MANDATORY)
70
71	Verify EVERY constraint from the DISCOVER phase `CONSTRAINTS CAPTURED` block
72	appears in the spec. If "must use Python" was a constraint, the spec MUST
73	reference Python. Missing constraints = spec bug.
74
75	Emit the check explicitly:
76
77	```
78	CONSTRAINT CHECK:
79	- Tech stack (Python): FOUND in spec section "Technical Stack"
80	- Timeline (MVP in 2 weeks): FOUND in spec section "Milestones"
81	- ...
82	```
83
84	Every constraint must be marked FOUND. If any are MISSING, loop back and
85	fix the spec before proceeding.
86
87	### SCOPE CHECK (MANDATORY)
88
89	Use the scale classification from DISCOVER to set task count range:
90
91	\| Scale \| Task Count \| Subtask Depth \|
92	\|------------\|-----------\|------------------\|
93	\| Solo \| 8–12 \| 2–3 subtasks each \|
94	\| Team \| 12–20 \| 3–5 subtasks each \|
95	\| Enterprise \| 20–30 \| 5–8 subtasks each \|
96
97	If DISCOVER classified the project as Team but the spec implies 30 tasks,
98	that's a scope bug — narrow the spec or re-classify explicitly.
99
100	### Domain-neutral vocabulary
101
102	When the domain is unclear, default to neutral terms:
103
104	\| Software term \| Neutral equivalent \| When to use neutral \|
105	\|---------------\|-------------------\|---------------------\|
106	\| tests \| verification criteria \| pentest, business, learning \|
107	\| code

People who liked this also viewed

6 picks

Type