Training · AutomatorMoveemergingverified

Shared Move Library

also known as prompt library, recipe library, prompt cookbook, corporate prompt library, AI playbook library

Craft Path: AutomatorMaker

A team-maintained, searchable collection of proven prompts and automation blueprints that anyone can reuse and improve. It democratizes AI expertise and locks in gains from Build Clinics and Build Sprints.

How the learner advances

Intent. Preserve and multiply the value of individual AI discoveries by making proven prompts and automation blueprints searchable and reusable across a team.

When to apply. Use when a team runs any recurring AI build activity — clinics, sprints, or individual experimentation — and wants the gains from each session to compound rather than stay siloed in one person's workflow. Most valuable when team members share similar tasks but reach different quality levels because they each prompt from scratch.

Threshold — earns the next step. A team member new to a task can find a relevant library entry, adapt it to their specific need without asking the original author for help, and produce a result that meets their quality bar on the first or second attempt.

Masterpiece — the artifact that proves it. A library entry that has been reused and adapted by at least three team members across different but related tasks, reducing their time-to-first-good-result compared to prompting from scratch — demonstrable by version history and usage notes.

Facets

  • Container — async
  • Mode — hands-on-buildpair-cohort
  • Reach — function
  • Persona — analyst-opsnon-technicalmanager-leaderbuilder
  • Craft (AI Fluency) — descriptiondiscernmentdiligence
  • Learner — human
  • Trainer — human
  • Guardrail — securityip-copyrightresponsible-use

Inputs

  • Proven prompts and automation blueprints — Tested, working prompts and workflow configurations produced during Build Clinics, Build Sprints, or individual practice — the raw material the library organises.
  • Shared platform with permissions — A searchable platform (Notion, SharePoint, TeamAI, PromptFluent, or equivalent) with folder categories, naming conventions, and role-based edit gates. Anyone can read and use; only reviewers can add or modify.
  • Review role assignment — At least one named person per team or department who vets new submissions for quality, accuracy, and security before they appear in the searchable library.

Outputs

  • A more capable learner — A team member who finds and adapts an existing library entry arrives at a working result faster and learns what good prompting looks like by reading tested examples, not just their own attempts.
  • Reusable prompt or blueprint (Masterpiece) — A documented, reviewed, production-tested entry in the shared library that another team member has successfully reused to solve their own task without asking the original author for help.
  • Organisational AI memory — A living body of team-specific AI knowledge that survives staff turnover and makes the organisation incrementally more capable each time a good prompt is added.

Steps (4)

  1. Choose platform and set structure

    Select a platform with search, categories, and role-based permissions. Agree on a naming convention (role + task + tool, e.g. 'ops-weekly-report-make') before the first entry is added; retrofitting structure onto a chaotic library is painful. Create an initial folder taxonomy by role or task type.

  2. Populate from existing events

    Immediately after every Build Clinic or Sprint, winning prompts and blueprints are deposited within 24 hours while context is fresh. The deposit window must be mandatory, not optional — the review role checks and approves within one business day.

  3. Version and maintain

    Every entry carries a version number and a last-tested date. The review role runs a quarterly audit: entries untested in the last 90 days are marked 'needs review'; entries failing a re-test are archived. Outdated entries that stay in the library quietly erode trust.

  4. Recurring demo ritual

    Run an optional 15-minute 'Prompt Friday' or equivalent slot where one team member demos a new or notably improved library entry to the group. Keeps the library visible, surfaces the best recent additions, and creates a lightweight social incentive to contribute quality entries.

Principles

  • A prompt that only one person knows is an asset at risk; a prompt in the library is organisational capability.
  • Quality gates beat quantity goals — 20 reviewed, working entries are more valuable than 200 untested ones.
  • Iteration is policy: the first library entry for any task is a starting point, not a finished product; version tracking makes improvement visible.
  • Outdated entries erode trust faster than an empty library; archive relentlessly.

Unlocks methodologies (2)

A learner who completes this pattern is equipped to execute these methodology families:

Prompt EngineeringIteration Management

Known uses (3)

Known failure modes (3)

  • [no-quality-gate]

    The anti-pattern of allowing anyone to add untested prompts without review. The library fills quickly but becomes unreliable; team members stop consulting it after two or three bad results.

  • [no-maintenance-ritual]

    Publishing entries without version tracking or a staleness audit. Prompt quality degrades silently as models update and workflows change; the anti-pattern creates a library full of entries that look authoritative but fail in practice.

  • [hoarding]

    The anti-pattern of keeping effective prompts private to protect individual productivity advantage. Common in competitive team cultures; it directly prevents the library from reaching the density needed to be useful.

Related trainings (2)

Sources (4)

Provenance

  • Ecosystem: neutral
  • Added to catalog:
  • Last updated:
  • Verification status: verified