This repository is a public documentation surface for Tiinex continuity, provenance, lineage, and recoverable markdown workflows.
The core idea is simple:
- preserve relationships between artifacts over time
- keep those relationships readable without requiring one specialized runtime
- separate
Parent Trace,Origin, schema identity, and integrity instead of flattening them into one vague metadata layer
- .topics/README.md: topic-oriented trace artifacts and continuity roots
- .topics/.schemas/README.md: human-readable schema notes for Tiinex artifact types such as topic, task, evidence, feedback, decision, pointer, and continuity envelope
Many provenance problems appear when relationships between observations, evidence, decisions, tasks, and later follow-up are preserved only inside one tool, one database, or one vendor-specific workflow.
Tiinex explores a markdown-first alternative where a reader can still inspect:
- what an artifact currently is
- which parent trace it directly continues
- what external or upstream origin grounded it
- which schema governs the current reading
- what integrity relation has been checked
Parent Trace: the direct continuity relation to the prior trace stepOrigin: the best known bounded reconstruction entrypoint or upstream grounding sourceLineage: the readable chain formed by trace artifacts over timeContinuity: the envelope and integrity discipline that help later readers continue work without guessingSchema: the artifact contract that tells a reader whether something is a topic, task, evidence, feedback, decision, pointer, runtime artifact, or another Tiinex type
If you are trying to understand Tiinex provenance or lineage thinking, start here:
- .topics/.schemas/tiinex.continuation.v1.md
- .topics/.schemas/tiinex.topic.v1.md
- .topics/.schemas/tiinex.task.v1.md
- .topics/.schemas/tiinex.evidence.v1.md
- .topics/.schemas/tiinex.decision.v1.md
- .topics/.schemas/tiinex.feedback.v1.md
- .topics/.schemas/tiinex.pointer.v1.md
These files are intentionally optimized for both humans and language models.
They are not meant to hide provenance inside opaque metadata. They are meant to keep parent, origin, schema, and continuity explicit enough that later readers can reconstruct what happened and why.