Prd V10 Continuous Discovery Torres

What it does

When your agent uses it

• Requests to set up discovery cadence
• Build opportunity solution tree
• Run weekly customer interviews
• User asks Torres

Example prompts

• “Torres”
• “continuous discovery”
• “opportunity solution tree”
• “/prd-v10-continuous-discovery-torres”

Requirements

• Pre-approved tools (allowed-tools): Read, Write, Edit, Glob, Grep

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

1. 1Define one measurable outcome — Not an output ("ship feature X"), but an outcome ("activated users in beachhead segment grow 20% MoM")…
2. 2Set up weekly cadence — 3+ customer interviews per week, ongoing. Not "until we feel done." Continuous.
3. 3Map opportunities under the outcome — Each opportunity is a customer pain or need (not a feature). Phrased in customer words. Grouped…
4. 4Pick top opportunity — Score by outcome-impact × evidence-strength × addressability. Focus on one at a time.
5. 5Brainstorm solutions — Multiple candidate solutions per opportunity. Not "the obvious one." Force divergent options.
6. 6Assumption-map the top solution — What must be true for this solution to work? Three categories: desirability (do they want it?)…
7. 7Test the riskiest assumption first — Smallest experiment that disproves the assumption if it's wrong. Update tree.

What it can do on your machine

Read from SKILL.md and the folder at commit 30ed1b0. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves these tools, so the agent can use them without asking each time:

  • Read
  • Write
  • Edit
  • Glob
  • Grep

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  No scripts in the folder and no shell commands in SKILL.md.

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Prd V10 Continuous Discovery Torres loads about 2.8k tokens when it runs. Until then it costs about 140 tokens; SKILL.md has 1,063 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~140

When it runs · the whole SKILL.md, loaded when a task matches

~2.8k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from mattgierhart/PRD-driven-context-engineering at commit 30ed1b0, republished under its MIT licence (© mattgierhart). 1,063 words, ~2,787 tokens.

name

prd-v10-continuous-discovery-torres

description

Establish weekly customer-discovery cadence and Opportunity Solution Tree practice using Teresa Torres's Continuous Discovery Habits framework during PRD v1.0 Market Adoption. Triggers on requests to set up discovery cadence, build opportunity solution tree, run weekly customer interviews, or when user asks "Torres", "continuous discovery", "opportunity solution tree", "outcomes vs outputs", "weekly interviews", "assumption mapping". Outputs CFD-* discovery entries and updates to ADO-STAGE-* and PER-* with new evidence.

allowed-tools

Read, Write, Edit, Glob, Grep

context

fork

execution_modes.default

standard

execution_modes.supports

quick, standard, deep

# Opportunity Solution Tree — [Date / EPIC]

## Outcome
KPI-XXX: [measurable outcome statement]

Anchored in: ADO-STAGE-AAA (stage assessment), ADO-BEACHHEAD-BBB (segment)
Time-bound: [timeframe]

## Opportunities

### O1: [Customer pain in their words]
- Frequency: [N interviews mention this]
- Confidence: X/5
- Outcome-link: [How does solving this move the outcome?]
- CFD-* sources: CFD-XXX, CFD-YYY
- Status: [Active | Deprioritized | Solved]

  #### Solutions for O1

  - S1.1: [Candidate solution]
    - Outcome-impact: [Predicted lift]
    - Effort: [Rough scope]
    - Status: [Brainstormed | Assumption-mapped | Experimenting | Validated | Killed]

    ##### Assumptions for S1.1
    - D1 [desirability]: [Must be true about user wanting it]
    - V1 [viability]: [Must be true about business impact]
    - F1 [feasibility]: [Must be true about building it]

    Test plan for [riskiest assumption]:
    - Experiment: [Smallest test]
    - Success threshold: [Specific metric]
    - Failure path: [What we do if it fails]
    - Status: [Planned | Running | Result]

CFD-XXX: Discovery Interview — [interview title]
Type: Discovery-Interview
Date: YYYY-MM-DD
Interviewee segment: [PER-XXX] [in-beachhead: yes/no]
Interviewer: [Name]

Key story (specific past behavior, not opinion):
  [Mom Test-disciplined quote — what they DID, not what they THINK]

Pain mentioned: [One concrete pain in their words]
Workaround used: [What they currently do]
Feature requests (discounted): [What they asked for — note as IDEA, not data]

Confidence: [3/5 — qualitative single interview; 4/5 — pattern across cohort]
Linked outcomes / opportunities: [KPI-XXX, O1, O3]
Tree position: [Which opportunity this evidence supports]

Linked IDs: PER-XXX, ADO-BEACHHEAD-XXX, KPI-XXX

© mattgierhart, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .claude/skills/prd-v10-continuous-discovery-torres of mattgierhart/PRD-driven-context-engineering.

Open the folder on GitHubat commit 30ed1b0