Task to PRD Generator
User-facing entry point for the task → PRD pipeline. The user runs
/task-to-prd and this skill dispatches to the companion babysitter
process via /babysitter:call (interactive — Five Whys, clarification
loop, scope-lock, per-finding gates, final approval) or
/babysitter:yolo (non-interactive, auto-approve every gate).
Usage
/task-to-prd <tracker-id | file-path | inline-text>
Examples:
/task-to-prd JIRA-1234
/task-to-prd path/to/email.txt
/task-to-prd "Eden asked to change the donut to show lessons" (inline)
Companion process
- Library path:
library/specializations/product-management/task-to-prd.js
- Locally installed:
~/.a5c/processes/task-to-prd.js
The process file is the source of truth for the pipeline. This skill
intentionally stays thin so users only need to remember /task-to-prd.
What this skill does when invoked
Parse input — detect whether it's a tracker ticket ID, a file
path, or inline text.
Infer featureBranch — read the current git branch via
git branch --show-current. If unsuitable (e.g., main), ask the
user to provide a feature-branch name.
Detect optional inputs from the project context:
contextDoc — project conventions doc (default CLAUDE.md)
archiveDir — past PRD archive directory (default empty)
trackerHint — hint like jira, linear, gh-issues (default empty)
secondaryReviewer — hint like codex, gemini, peer (default empty)
Ask the user to pick the run mode:
Run mode for the babysitter task-to-prd process:
- Interactive (
/babysitter:call) — recommended. The Five
Whys + clarification loop is intrinsically interactive — every
question goes through the user. Verification findings each get
a per-finding approval gate.
- Auto / yolo (
/babysitter:yolo) — non-interactive. Skips
the clarification loop (uses defaults from Five Whys), auto-
approves every gate, applies all proposed PRD changes
automatically. Use only when the input is already well-defined
and you want a fast first draft.
Default to option 1 if the user does not respond.
Dispatch by invoking the chosen babysitter command via the Skill tool:
For interactive:
Skill('babysitter:call', 'Run the product-management/task-to-prd process with inputs: input=<resolved>, featureBranch=<resolved>, contextDoc=<resolved>, archiveDir=<resolved>, trackerHint=<resolved>, secondaryReviewer=<resolved>')
For auto / yolo:
Skill('babysitter:yolo', '<same instruction as above>')
Return the PRD path, the Decision Log, and the follow-up
/prd-to-spec prompt produced by the process to the user.
Fallback (no babysitter runtime)
If neither /babysitter:call nor /babysitter:yolo is available:
- Read the process source at the location above.
- Execute each task's agent prompt manually, in phase order.
- Pause at every
ctx.breakpoint(...) for explicit user approval
(interactive equivalent) or skip every breakpoint (yolo equivalent),
matching the mode the user picked in step 4.
- Return the same outputs the process would have returned.
Optional Integration Hooks
All flow through the process inputs (none required):
- Tracker integration —
trackerHint input. Examples: Jira
(/jira skill or acli jira workitem view), Linear (CLI / MCP),
GitHub Issues (gh issue view).
- Secondary reviewer —
secondaryReviewer input. Examples: Codex
(/codex:review), Gemini (/gemini-review), deep-verify-plan
(/deep-verify-plan), peer (notify a teammate).
- Stakeholder attribution — the Decision Log records each Q&A with
attribution to
user or a named stakeholder (e.g., product owner,
tech lead, designer).
The process works end-to-end without any of them.
Pipeline shape (executed by the process)
| Phase |
What happens |
Interactive in /babysitter:call? |
| 1a. Load source |
Detect tracker / file / inline |
no |
| 1b. Five Whys |
Root-cause analysis |
no |
| 1c. Clarification loop |
Q&A one at a time |
YES |
| 1d. Scope Lock |
Approve scope before drafting |
YES |
| 2. Draft (parallel) |
Codebase scan + draft PRD via embedded skill instructions |
no |
| 3a. Verification (5 parallel) |
what-could-go-wrong + consistency + conventions + adversarial + quality checklist |
no |
| 3b. Per-finding gate |
Approve each proposed PRD change |
YES |
| 4a. Final review |
Approve final PRD |
YES |
| 4b. Tracker update |
Push reference to tracker (if applicable) |
no |
| 4c. Follow-up prompt |
Output /prd-to-spec <path> |
no |
In /babysitter:yolo mode, every gate marked YES is auto-approved.
Rules
- The skill itself does not produce the PRD. It only sets up inputs
and dispatches to the process via
/babysitter:call or /babysitter:yolo.
- Default to
/babysitter:call (interactive) when in doubt.
- Phase 1 (clarification) is the most user-intensive — yolo mode
effectively turns it into a shallow first-draft using only Five-Whys
defaults. Recommend interactive mode for any non-trivial task.
- The PRD must be proportional to the task size (the process honors this).
1---
2name: task-to-prd
3description: "Convert a raw task (tracker ticket, email, or text) into a fully characterized PRD. User-facing entry point that dispatches to the task-to-prd babysitter process via /babysitter:call (interactive) or /babysitter:yolo (auto). Stack-agnostic."
4author: Yehuda Yungstein
5graph:
6 domains: [domain:software-engineering]
7 specializations: [specialization:product-management]
8 skillAreas: [skill-area:product-discovery, skill-area:roadmap-planning]
9 roles: [role:product-manager, role:product-analyst]
10 workflows: [workflow:product-discovery, workflow:competitive-analysis]
11metadata:
12 source: https://skillmd.com/skills/a5c-ai/task-to-prd
13---
14
15# Task to PRD Generator
16
17User-facing entry point for the task → PRD pipeline. The user runs
18`/task-to-prd` and this skill dispatches to the companion babysitter
19process via `/babysitter:call` (interactive — Five Whys, clarification
20loop, scope-lock, per-finding gates, final approval) or
21`/babysitter:yolo` (non-interactive, auto-approve every gate).
22
23## Usage
24
25```
26/task-to-prd <tracker-id | file-path | inline-text>
27```
28
29Examples:
30- `/task-to-prd JIRA-1234`
31- `/task-to-prd path/to/email.txt`
32- `/task-to-prd "Eden asked to change the donut to show lessons"` (inline)
33
34## Companion process
35
36- Library path: `library/specializations/product-management/task-to-prd.js`
37- Locally installed: `~/.a5c/processes/task-to-prd.js`
38
39The process file is the source of truth for the pipeline. This skill
40intentionally stays thin so users only need to remember `/task-to-prd`.
41
42## What this skill does when invoked
43
441. **Parse input** — detect whether it's a tracker ticket ID, a file
45 path, or inline text.
462. **Infer `featureBranch`** — read the current git branch via
47 `git branch --show-current`. If unsuitable (e.g., `main`), ask the
48 user to provide a feature-branch name.
493. **Detect optional inputs** from the project context:
50 - `contextDoc` — project conventions doc (default `CLAUDE.md`)
51 - `archiveDir` — past PRD archive directory (default empty)
52 - `trackerHint` — hint like `jira`, `linear`, `gh-issues` (default empty)
53 - `secondaryReviewer` — hint like `codex`, `gemini`, `peer` (default empty)
544. **Ask the user to pick the run mode:**
55
56 > Run mode for the babysitter `task-to-prd` process:
57 >
58 > 1. **Interactive (`/babysitter:call`)** — recommended. The Five
59 > Whys + clarification loop is intrinsically interactive — every
60 > question goes through the user. Verification findings each get
61 > a per-finding approval gate.
62 > 2. **Auto / yolo (`/babysitter:yolo`)** — non-interactive. Skips
63 > the clarification loop (uses defaults from Five Whys), auto-
64 > approves every gate, applies all proposed PRD changes
65 > automatically. Use only when the input is already well-defined
66 > and you want a fast first draft.
67
68 Default to option 1 if the user does not respond.
69
705. **Dispatch** by invoking the chosen babysitter command via the Skill tool:
71
72 For interactive:
73 ```
74 Skill('babysitter:call', 'Run the product-management/task-to-prd process with inputs: input=<resolved>, featureBranch=<resolved>, contextDoc=<resolved>, archiveDir=<resolved>, trackerHint=<resolved>, secondaryReviewer=<resolved>')
75 ```
76
77 For auto / yolo:
78 ```
79 Skill('babysitter:yolo', '<same instruction as above>')
80 ```
81
826. **Return** the PRD path, the Decision Log, and the follow-up
83 `/prd-to-spec` prompt produced by the process to the user.
84
85## Fallback (no babysitter runtime)
86
87If neither `/babysitter:call` nor `/babysitter:yolo` is available:
88
891. Read the process source at the location above.
902. Execute each task's agent prompt manually, in phase order.
913. Pause at every `ctx.breakpoint(...)` for explicit user approval
92 (interactive equivalent) or skip every breakpoint (yolo equivalent),
93 matching the mode the user picked in step 4.
944. Return the same outputs the process would have returned.
95
96## Optional Integration Hooks
97
98All flow through the process inputs (none required):
99
100- **Tracker integration** — `trackerHint` input. Examples: Jira
101 (`/jira` skill or `acli jira workitem view`), Linear (CLI / MCP),
102 GitHub Issues (`gh issue view`).
103- **Secondary reviewer** — `secondaryReviewer` input. Examples: Codex
104 (`/codex:review`), Gemini (`/gemini-review`), `deep-verify-plan`
105 (`/deep-verify-plan`), `peer` (notify a teammate).
106- **Stakeholder attribution** — the Decision Log records each Q&A with
107 attribution to `user` or a named stakeholder (e.g., product owner,
108 tech lead, designer).
109
110The process works end-to-end without any of them.
111
112## Pipeline shape (executed by the process)
113
114| Phase | What happens | Interactive in `/babysitter:call`? |
115|-------|--------------|-----------------------------------|
116| 1a. Load source | Detect tracker / file / inline | no |
117| 1b. Five Whys | Root-cause analysis | no |
118| 1c. Clarification loop | Q&A one at a time | YES |
119| 1d. Scope Lock | Approve scope before drafting | YES |
120| 2. Draft (parallel) | Codebase scan + draft PRD via embedded skill instructions | no |
121| 3a. Verification (5 parallel) | what-could-go-wrong + consistency + conventions + adversarial + quality checklist | no |
122| 3b. Per-finding gate | Approve each proposed PRD change | YES |
123| 4a. Final review | Approve final PRD | YES |
124| 4b. Tracker update | Push reference to tracker (if applicable) | no |
125| 4c. Follow-up prompt | Output `/prd-to-spec <path>` | no |
126
127In `/babysitter:yolo` mode, every gate marked YES is auto-approved.
128
129## Rules
130
131- The skill itself does not produce the PRD. It only sets up inputs
132 and dispatches to the process via `/babysitter:call` or `/babysitter:yolo`.
133- Default to `/babysitter:call` (interactive) when in doubt.
134- Phase 1 (clarification) is the most user-intensive — yolo mode
135 effectively turns it into a shallow first-draft using only Five-Whys
136 defaults. Recommend interactive mode for any non-trivial task.
137- The PRD must be proportional to the task size (the process honors this).