More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Product Manager Toolkit: RICE, Interviews and PRDs

    1

    One connected system for the PM loops: RICE scoring with portfolio balance, interview synthesis that produces roadmap evidence, and right-sized PRDs with agreed acceptance criteria.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Product Manager Toolkit: RICE, Interviews and PRDs

    Product Manager Toolkit: RICE, Interviews and PRDs

    Example session with this skill installed

    Plan our quarter. We have six feature candidates, a pile of customer interview notes about onboarding, and a team of four engineers. Tell me what we build, in what order, and spec the top item.

    • Read your context and instructions
    • Compiled the product manager toolkit

    QUARTER PLAN — produced by applying the skill.

    Learn first (the evidence): eight onboarding interviews synthesized. Dominant theme: six of eight users stalled connecting their data source; five raised it unprompted. Two verbatim quotes anchor the theme, including "I signed up on Monday and didn't get a single chart until Thursday." Feature requests from interviews are captured but tagged as requests, not requirements.

    Decide (RICE + portfolio): Guided data-source setup — Reach 2,000/quarter × Impact 2 × Confidence 80% (interview-backed) ÷ Effort 1.5 = 2,133, top score, and a quick win. SSO — 890, lower score but strategically required: selected as the quarter's single big bet, with the override and its reason written on the plan. Two fill-ins (empty-state guidance, invite-flow copy fix) fit around the bet. The reporting rebuild scores 210 at high effort: money pit, killed in the plan document with its score shown, so it stays killed in week six when someone asks again. Total selected effort: 8.5 person-months against 9.5 real capacity. Fits.

    Specify (top item): one-pager PRD for guided setup — problem with the two quotes; success metric setup completion 45% → 70%, guardrail that setup-related support tickets must not rise; out of scope: custom connectors; acceptance criteria covering the happy path and the three error states the interviews surfaced. Engineering pre-review signed off before the planning meeting, so the meeting argued priorities, not the problem statement.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Product work fragments when its three loops run apart: the backlog gets ranked by whoever argues loudest, interview notes sit in a folder nobody cites, and PRDs describe everything with equal urgency. This skill connects the loops so each one feeds the next. Loop one decides: every candidate scored with RICE — Reach, Impact, Confidence, Effort — where Confidence is treated as the honesty knob and scoring disagreements are surfaced as information rather than averaged away. The ranked list then faces the step most teams skip: a portfolio balance check (quick wins, at most two big bets, fill-ins, and money pits killed in writing) mapped against the team's real capacity, not its nominal headcount math. Loop two learns: a semi-structured interview method built on past behavior instead of hypotheticals, and a synthesis discipline that extracts pains with severity, jobs to be done, feature requests tagged as requests, and the verbatim quotes that actually persuade stakeholders — then clusters themes across interviews so the loudest pain becomes cited evidence that raises Confidence scores in loop one. Loop three specifies: PRDs sized to the work (one-pager, standard, epic brief), a six-section structure whose most valuable part is the explicit out-of-scope list, acceptance criteria agreed with engineering before build starts, and a pre-review pattern that keeps the big review about priorities. A worked example runs a full quarter through all three loops. Includes a RICE worksheet, a PRD readiness rubric, and a quarterly planning checklist.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Productivity.

    • RICE makes assumptions explicit; it cannot make them true. Scores are only as good as the Reach and Effort estimates behind them, which is why the skill forces every score to carry its source or admit it is a guess.
    • It cannot manufacture user evidence. With no interviews or usage data, Confidence caps at low by design — the honest output is a hypothesis list and a discovery plan, not a confident roadmap.
    • It covers prioritization, discovery synthesis, and specification. Delivery management — sprint planning, estimation ceremonies, release coordination — is a different discipline and out of scope.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 24 days ago

    Permissions

    No special permissions declared or detected

    Listed24 days ago
    Updated21 days ago

    What's inside

    Frequently Asked Questions

    It can be, when the rank is treated as the answer. This skill treats the score as the start of a conversation: every factor must carry its evidence or admit it is a guess, disagreements between scorers are surfaced rather than averaged, and the ranked list still has to survive a portfolio balance and capacity check before anything is called a plan.

    Yes, with honest labels. The skill accepts degraded evidence — support tickets, sales call notes, churn reasons — and caps Confidence scores to match. What it will not do is let a plan built on zero user contact present itself as evidence-based; the synthesis loop then becomes a short discovery plan to run before the next quarter.

    Two ways. Size is chosen deliberately — one-pager, standard, or epic brief — because an oversized PRD gets its document reviewed instead of its problem understood. And the problem section must quote real user evidence from the interview loop, while the acceptance criteria (including error states) are agreed with engineering before build starts, not written afterward to describe whatever got built.

    New in Productivity