Skip to content

TAPE-AUDIT.md

Latest commit

 

History

History
24 lines (17 loc) · 1.79 KB

TAPE-AUDIT.md

File metadata and controls

24 lines (17 loc) · 1.79 KB

TAPE-AUDIT — void

Audit-class survey for .tape adoption (typed events + provenance edges + delivery grade).

A. Audit-class ledgers

CARGO only. state/markers/install_*.marker are repetitive install-trace markers (zero design content — pure install cargo). state/proposals/ may carry RFC-class proposals but no .jsonl event streams. No audit/ dir. No structured per-event history.

B. Identity surface

None — void is a terminal emulator (Ghostty fork, Zig + Swift). The "identity" surface is the user's running terminal session, but void doesn't track or model agent / system identity in any persistent form. Session state lives in the OS-level terminal pty.

C. Domain.md files

Heavy surface but documentation-class, not event-class: AI_POLICY.md, CHANGELOG.md, CONTRIBUTING.md, HACKING.md, LATTICE_POLICY.md, LIMIT_BREAKTHROUGH.md, PACKAGING.md, VOID_FORK.md, plus AGENTS.md / CLAUDE.md / README.md. These are governance / docs SSOTs — they fit the <UPPERCASE>.md convention but their content is policy text, not event data. Not natural .tape placement targets.

D. Per-run / per-event history

None. No benchmark / simulation / training ledgers. The product is a terminal emulator; runtime events are ephemeral keypresses + pty bytes.

E. Promotion candidates

  • n6 atoms: not applicable (no measurement facts).
  • .tape events: not applicable (no typed event surface beyond install markers, which are cargo).
  • hxc wire: arguably the natural byte wire IS the terminal itself, but that's .hxc philosophically, not a void adoption story.
  • n12 cube: not applicable.

Verdict

NONE — void is a pure code repo (terminal emulator). Markers are install cargo, doc files are governance text, no event-class ledgers. No .tape adoption surface.