Modules
Six optional modules, each graduated into its own spec. Install any subset.
Six graduated modules + the kit plugin. Zero inter-plugin dependencies.
Modules
Project Seed
A one-page starter that tells your AI assistant how to behave on your project.
Six-tier bootstrap. Phase 0 setup + AGENTS.md scaffold + coordination patterns.
README →Superscalar
Parallel AI work that stays consistent. Measured: 2.2× more grounded · 40% less guesswork.
Discipline for parallel sub-agent dispatch. issue_width + worktree isolation + retire + consistency gate.
Read the A/B report →Hyperbrief
A structured brief on every AI-to-user decision request: what / how / why / consequences / decision tree / recommendation. One-click re-render in the reader's plain language (button auto-localizes to user's language).
Decision-delegation gating discipline. 4-score escalation + JSON IR + deterministic renderer + 3-axis × 5-level tone profile. Byte-identical output verified. Integrated with Constellation + Superscalar.
Read the discipline →Greatpractice
Memory notes are the trigger; recurring patterns get promoted into auto-enforced practices. Backed by 9-axis cross-domain research.
Memory feedback → 5-axis maturation gate → ratified entry in 3-tier tree, enforced by 7-event lifecycle hooks. phronesis_boundary protects un-codifiable work.
Read the discipline →Ultrasafe
Several AI helpers find weak spots before release, fix-and-retry at least three rounds until clean.
Pre-release simulated penetration testing. 8-agent fan-out + ≥3 iteration AND-gate + 3-layer synthesis. Advisory→blocking.
Read the discipline →Constellation
Multiple AI agents talking in real time + shared dashboard + safe handoffs. Five real PRs in 24h, zero mistakes.
Live multi-agent orchestration: WebSocket + A2A + dashboard + MCP. 24h dogfood: 5 e2e PRs, 0 violations. + v2.4.24–46: PWA/Web Push, access control, wiki side panel, ack ✓, echo-mode.
Read the protocol →Compendium
A portable vocabulary discipline over any vault + dual-register glossary (internal ↔ general term) with click-to-define cross-links. Owns nothing normative — it points into the module specs.
Concept-anchored dual-register vocabulary substrate: wiki + glossary mapping internal ↔ general term, cross-link / click-to-define. Owns nothing normative (pointers into module specs). The north-star axis-2 vocabulary-substrate module.
Read the module →The skills — what you can actually call
Agent Skills — the full roster (33)
A skill is a small procedure document the agent reads and follows. Below is everything the seven plugins bring, grouped by plugin.
36 SKILL.md files across the 7 plugins, three invocation kinds: user slash commands, model-invoked, and orchestrator-dispatched. The same files project onto Codex via the adapter.
estregenesis 6 procedures + 5 aliases
/eg-bootstrap(alias/egboot) — set a new project up on the EG standard./eg-migration(alias/egmig) — move a project that already has its own rule files./eg-upgrade(alias/egup) — bring a seeded project to the current release./eg-memsync(alias/egmem) — reconcile agent memory with the governance docs./egrich— full arm: the seed plus all six modules, wired and verified./eg-interview(alias/eggrill) — interview a request into a delegable brief.
constellation 9 skills
/roundtable— conduct discipline for multi-party topic rooms./boardsweep— true-up every agent-authored board entry against real repo history./roundnext— one what-next round over the live board./drillnow— attack the top blocker right now./before-compact— materialize volatile session details before compaction./echo-mode— mirror local chat onto the board channel./constellation-start— start the live board (user command).constellation-board(model-invoked) — read live board state.constellation-a2a-emit(model-invoked) — send a targeted A2A message.
hyperbrief 3 skills, model-invoked
hyperbrief-trigger-check— cheap rubric before any decision-ask.hyperbrief— generate the decision-delegation brief.hyperbrief-revisit— revisit a stored decision and record the outcome.
superscalar 2 skills
superscalar— consulted before any parallel sub-agent dispatch./subscaler— toggle tiered model composition.
greatpractice 1 skill
/routinize— capture a recurring work pattern into durable practice.
ultrasafe 8 skills, orchestrator-dispatched
Seven attacker perspectives probing a release in parallel, plus one synthesizer merging their findings.
ai-llm-redteam·web-api-attacker·supply-chain-auditor·crypto-reviewersocial-engineer·methodology-compliance·threat-model-lifecyclesynthesizer— aggregates all seven attackers' findings. Advisory in v0.2.x.
compendium 2 skills, cross-tool
compendium-curate— add or maintain a vocabulary entry.compendium-lint— health-check the vocabulary store.