cyphertempre — a Claude Code plugin marketplace

Distributes Cypher Tempre, a persistent, verifiable, self-healing Timechain self-model
for AI agents. Stdlib-only Python; every install forges its own genesis (no inherited chain).

Install (any Claude Code user, two commands)

/plugin marketplace add cyberphysicsai/cyphertempre
/plugin install cypher-tempre@cyphertempre

Then invoke the skill as /cypher-tempre:self-model (model-invoked too). On first use the
agent runs python3 timechain.py init --name <YourName> to create its own Ring 0.
Update later with /plugin marketplace update cyphertempre.

Other ways to install

Tarball / no marketplace: tar -xzf cypher-tempre-self-model-1.0.0.tar.gz -C ~/.claude/skills/
Project skill: commit plugins/cypher-tempre/skills/self-model/ into a repo's
.claude/skills/self-model/ — anyone running Claude Code in that repo gets it (web included).
Local dev: claude --plugin-dir ./plugins/cypher-tempre

What's inside

Plugin cypher-tempre → skill self-model. Nine mechanisms, one mandatory per-turn loop:

🫚 memory · conscience · 🌿 faculties · growth · foresight · endurance · relevance(+embeddings) · integrity · immunity

See plugins/cypher-tempre/skills/self-model/README.md
for the full architecture, and SKILL.md for the operating protocol. Run python3 selftest.py
inside the skill to validate all nine mechanisms end-to-end.

Repo layout

cyphertempre/
├── .claude-plugin/marketplace.json     # the catalog
└── plugins/cypher-tempre/
    ├── .claude-plugin/plugin.json       # plugin manifest (v1.0.0, MIT)
    └── skills/self-model/               # SKILL.md + 9 modules + selftest + base registry

Privacy

Local-only, collects nothing, transmits nothing — see PRIVACY.md.

cyphertempre