aethermind-primitive

agent
Security Audit
Warn
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 5 GitHub stars
Code Pass
  • Code scan — Scanned 12 files during light audit, no dangerous patterns found
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

Universal harness-free AetherMind continuity primitive for agent work.

README.md

AetherMind

AetherMind is an open source continuity primitive for AI-assisted work.

It preserves compact working judgment beside the project it describes: decisions,
corrections, discoveries, uncertainty, friction, verification, and the relationships
that make those records current or historical.

Install

python -m pip install .

The package installs the aethermind command and the aethermind Python module.

Quickstart

The original 0.1 commands remain available:

aethermind init --root . --purpose "start project continuity"
aethermind layer --root . --type decision --ctx build --body "Keep continuity beside the project"
aethermind validate --root .
aethermind inspect --root .

The 0.2 command names expose the complete local runtime:

aethermind write-layer --project-root . --type discovery --ctx build --body "The adapter path is working"
aethermind currentness --project-root .
aethermind brief --project-root .
aethermind capabilities --project-root .

Store layout

AetherMind continuity uses .aem files inside the project-local
.aethermind/ directory:

.aethermind/
  layers.aem
  texture.aem
  events.aem
  archive.aem

layers.aem is the decision ledger. texture.aem holds compact attention
pointers. events.aem keeps routine observations separate from the decision
ledger. archive.aem preserves archived record copies; layers.aem receives
the append-only tombstone that makes the archived records historical.

What changed in 0.2

AetherMind 0.2 extends the original primitive with the following continuity
methods:

  • six supported primitives: layer, artifact-reference, anchor,
    pressure-event, supersession, and rollback;
  • explicit store initialization and acknowledged append results;
  • filtered reads, currentness projection, scoped briefs, and anchor briefs;
  • append-only texture, event, and archive stores;
  • correction, supersession, rollback, evidence, recurrence, verification, and
    artifact relationships;
  • the original 0.1 Python helpers and CLI command names.

Existing 0.1 layers.aem files remain readable. New records append without
rewriting existing records. See Upgrading to 0.2.

Python

The 0.1 root-first API remains available:

from aethermind import init_store, read_layers, validate_store, write_layer

init_store(".", purpose="start project continuity")
write_layer(
    ".",
    type="decision",
    ctx="build/runtime",
    body="Use the project-local AEM store.",
    markers=["!"],
)
layers = read_layers(".")
status = validate_store(".")

The current engine is available through AetherMind:

from aethermind import AetherMind

store = AetherMind(".", create=False)
brief = store.brief()
current = store.currentness()

Remote work

aethermind remote-note writes continuity to a local remote-work store while
recording the target as metadata:

aethermind remote-note \
  --remote example-host:/srv/app \
  --task incident-1 \
  --ctx triage \
  --body "Inspected the service and retained the decision locally"

Documentation

Reviews (0)

No results found