You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[Submission] Tin — durable marketing workflows for coding agents #153
An Apache-2.0 marketing system for coding agents, using Temporal’s Python SDK to coordinate versioned workflows, retries, schedules, human review and isolated E2B execution.
Long Description
Tin (https://tin.computer/) turns repeatable marketing work into versioned workflows that developers can run from a dashboard or coding agent over MCP. The open-source application covers SEO research, content, outreach and product QA.
Temporal is the durable orchestration layer: Python workflows coordinate activities, timers, retries and waits for human review. Activities perform external I/O and sandbox execution; histories carry identifiers and small control values rather than prompts, file bodies or credentials. Postgres holds the product-facing state, and project git repositories retain outputs after E2B sandboxes are removed.
Useful examples for Temporal developers include long-running agent work with human approval, scheduled dispatch, immutable run/version pins, identifier-only history, and idempotent external-effect handling.
License: Apache-2.0. Self-hosting currently requires engineering setup and managed-service configuration; clean-clone turnkey-install acceptance remains outstanding, as documented in the repository.
Author(s)
Tin Computer — https://tin.computer/
Built by Ege Özgirin and Emre Şarbak; maintained by the Tin Computer team.
Activity
github-actions commentedon Nov 11, 2026
ZiggyBot — automated pre-evaluation
An error occurred during automated evaluation; the community team will need to review this submission manually.
Automated AI pre-screen by ZiggyBot. Final decisions are made by the community team.
github-actions commentedon Nov 11, 2026
ZiggyBot — automated pre-evaluation
An error occurred during automated evaluation; the community team will need to review this submission manually.
Automated AI pre-screen by ZiggyBot. Final decisions are made by the community team.