Tink
Stop losing context between Claude Code and Codex runs.
Tink keeps every non-trivial agent task in visible files - a task contract, run state,
verification steps, and reusable harnesses that are saved only after your approval.
No server, no telemetry, no hidden state.
A small harness layer for Claude Code and Codex
Latest package: v1.12.0 - Tink now records a human-readable evidence card after verification, adds the strict completion-policy groundwork for "no evidence, no done", and shows dashboard hints for harness trust, ROI, and failed/blocked run review. See CHANGELOG for release history.
English · 한국어 · Changelog
Who this is for
Use Tink when:
- Claude Code or Codex keeps losing task context between runs
- you repeat the same review / refactor / debug workflow by hand
- you want visible run state instead of hidden chat memory
- you want reusable agent workflows - saved only after explicit approval
If that sounds like your day, try it on a throwaway repo first:
npx tink-harness@latest install
What you actually get
Every non-trivial task leaves plain files you can open, diff, and commit:
.tink/current/ # the active run - always inspectable
contract.json # what must be true when the task is done
plan.md # the visible plan
checks.md # verification to run before claiming "done"
.tink/runs/
2026-06-11-1430-auth-refactor.md # compact record of each finished run
.tink/harnesses/
refactor-review.md # reusable ways of working - approval-gated
Why not just CLAUDE.md / slash commands / skills?
| Tooling | What it gives you | What Tink adds on top |
|---|
| CLAUDE.md | project-wide instructions | per-task contracts, run state, and verification |
| Slash commands | reusable prompts | harness selection, run records, progress tracking |
| Skills | reusable capability | usage lifecycle: health scores, cleanup and improvement signals |
| MCP | external context and tools | local, approval-gated workflow memory |
Tink composes with all of these - it does not replace them.
Install & quick start
Try Tink in about a minute.
Claude Code (plugin):
/plugin marketplace add dotoricode/tink-harness
/plugin install tink@tink-harness
/reload-plugins
/tink:setup
Claude Code or Codex (standalone):
npx tink-harness@latest install
The standalone installer auto-detects LANG (English fallback); pass --lang=en|ko|zh to override. During install you can pick Claude Code, Codex, or both - in Codex, start with $tink:cast <task>.
Repo-local Codex smoke test (CODEX_HOME)
set CODEX_HOME=%CD%/.codex
npx tink-harness@latest install --yes
Then hand Tink a real task instead of reading more docs:
/tink:cast refactor the auth module # Claude Code
$tink:cast refactor the auth module # Codex
cast picks (or drafts) the right harness, writes a visible plan into .tink/current/, and starts the first safe step after your approval. Every finished run leaves a compact record - and those records become the dashboard below.
See your harness health
After a few runs, one command turns your records into a local dashboard and opens it in your browser:
npx tink-harness dashboard # add --no-open to just generate the file
Under the hood it runs the two read-only helpers (node .tink/tools/generate-harness-lifecycle-summary.mjs, then node .tink/tools/render-harness-health-report.mjs) and opens .tink/maintenance/harness-health-report.html.
