AI Development Methodology

tinker — Strategy Master Plan

Fictional adopter strategy doc. Demonstrates the 01_strategy.md skeleton.

Vision (one sentence)

A command-line tool that makes capturing developer notes as low-friction as a git commit — so context that would normally be lost gets preserved, retrievable, and tied to the work it described.

Audience

Primary: solo developers who already use the command line for most of their workflow. They want a tool that fits where they already are, not another app to switch to.

Anti-audience: team-collaboration users who want shared notes with permissions, threading, and a web UI. tinker is single-user only by design; team-shared notes is a different product.

Business outcomes

• Free, open-source, MIT.
• Funded by maintainer’s other work; no monetization plan in v1.
• Success = adoption count, not revenue. KPI: 1,000 weekly active installs by 6 months post-public-launch.

Phases

Phase 1 — Foundation (current → 3 months)

Goal: working CLI with the core capture + retrieval flow.

Exit criteria (binary):

• tinker capture writes a note to local store; survives reboot.
• tinker recent retrieves the last N notes for the current directory.
• tinker search <query> returns matching notes ranked by recency.
• Install is one command on macOS, Linux, Windows (3 OS minimum).
• Test suite covers capture + retrieval + search; CI green on all 3 OS.

Phase 2 — Polish (3–6 months)

Goal: ship to closed beta of trusted developers; collect feedback; reach open-beta-ready quality.

Exit criteria (binary):

• Calendar-event integration works on macOS (calendar API). Windows + Linux integrations deferred to Phase 3.
• Notes link to git commits when in a repo.
• Closed beta of ≥ 5 trusted developers; ≥ 80% would recommend.
• Crash rate < 0.5% over a 30-day window.
• Milestone deep-eval scores ≥ 9 average across all areas (UX, perf, content quality, docs, test coverage, security).

Phase 3 — Open beta (6–9 months)

Goal: public on package managers; gather organic adoption signal.

Exit criteria (binary):

• Available on Homebrew, Apt, Cargo, Chocolatey.
• One blog post + Show HN.
• 100 weekly-active installs measured.
• ≥ 3 community contributors landed.

Phase 4 — v1.0 (12 months target)

Goal: declare v1 (drop “beta” label); commit to stability + support cadence.

Exit criteria (binary):

• All P0/P1 items from beta closed.
• Documentation complete; setup time < 5 minutes for a new user.
• 500 weekly-active installs.
• Milestone deep-eval scores ≥ 9 average, no area below 8.

Document index

#	Doc	Purpose
00	Master plan	This file — vision, phases, doc index.
01	(not yet authored) Market research	Survey of existing developer-notes tools; differentiation.
02	(not yet authored) Tech stack	Language choice (Rust), distribution mechanics.
03	(not yet authored) Privacy + data handling	Where notes live; what gets sent off-device (nothing in v1).

Re-evaluation

Cadence: at every phase exit + when a Phase exit criterion isn’t met within 50% of the target window.

Version history

Version	Date	Notes
v1.0	(fictional date)	Initial strategy doc.

This site is open source. Improve this page.