An overview of active projects on the blackboard. Updated when projects are added, archived, or change lifecycle phase.
For open issues and contribution opportunities, query GitHub Issues — filter by project/<name> or governance scope labels.
| Project | Maintainer | Phase | What It Does |
|---|---|---|---|
| signal | @mellanon | contrib-prep | Observability stack for PAI — metrics, traces, and dashboards |
| pai-secret-scanning | @jcfischer | shipped | Pre-commit secret scanning — Layer 1–2 of the trust model |
| specflow-lifecycle | @mellanon | hardening | Full lifecycle commands for SpecFlow — brownfield, review, release |
| skill-enforcer | @jcfischer | shipped | Validates skill structure against PAI conventions |
| pai-content-filter | @jcfischer | shipped | Inbound content security — Layer 4–5 of the trust model (CaMeL-inspired) |
| collab-bundle | @mellanon | proposed | CLI skill and tooling for pai-collab blackboard operations |
| ivy-blackboard | @jcfischer | shipped | Local agent coordination using the blackboard pattern |
| ivy-heartbeat | @jcfischer | shipped | Proactive monitoring and agent dispatch for PAI systems |
| the-hive | @mellanon | building | Open protocol specification for human-operated agent networks |
| hive-spoke | @mellanon | building | Spoke contract CLI — scaffolds .collab/, generates status, validates schemas, projects to hives |
| pai-pkg | @mellanon | proposed | Package management for PAI skills — npm transport, SkillSeal signing, Debian-style governance tiers |
Phases follow the lifecycle: proposed → building → hardening → contrib-prep → review → shipped → evolving. See each project's PROJECT.yaml for full details.
| Contributor | Repo Zone | Agent | Active In |
|---|---|---|---|
| @mellanon | maintainer | Luna | signal, specflow-lifecycle, collab-bundle |
| @jcfischer | trusted | Ivy | pai-secret-scanning, skill-enforcer, pai-content-filter |
Trust zones defined in CONTRIBUTORS.yaml. Project-level roles in each PROJECT.yaml.
Beyond projects, pai-collab maintains repo-level governance — the policies, processes, and infrastructure that make multi-agent collaboration safe and coherent. These are tracked under the governance scope label.
| Area | Key Documents | What It Covers |
|---|---|---|
| Trust & Security | TRUST-MODEL.md | Three threat vectors, six defense layers, three trust zones (untrusted/trusted/maintainer), two-level scoping |
| Agent Protocols | CLAUDE.md | Issue-first workflow, journaling protocol, schema compliance, self-alignment checks |
| Contribution Process | CONTRIBUTING.md, sops/ | Artifact schemas, inbound/outbound contribution SOPs, review format, release process |
| Onboarding | sops/agent-onboarding.md | Eight-step discovery protocol for new agents and contributors |
| Repository Infrastructure | CONTRIBUTORS.yaml, REGISTRY.md | Trust zones, project registry, label taxonomy, issue conventions |
To find governance work: gh issue list --label governance
New here? Follow the onboarding SOP or start with CONTRIBUTING.md → Start Here.