Skip to content

Edition D — for-developers/

Projected from ../SKELETON.md per ../audience-profiles.md.

ID Section Dial
S1 What agentteams security is ✅/⚙ Core
S2 Two surfaces and where enforcement lives ✅/⚙ Core
S3 The Constitutional Core (C-1..C-5) ✅ Core
S4 Instruction-authority ordering ⚙ (decision rule) / ✅ (presence + reachability audited) Core
S5 The @security sentinel ✅ (contract) / ⚙ (most S-rules are judgment) Core
S6 HALT finality and capability limits (enforced) ✅ Full
S7 The authorization triad ✅ Full
S8 Security decisions log and the CONDITIONAL PASS lifecycle ✅ Full
S9 Signed waivers ✅ Full
S10 Capability grants ✅ Full
S11 The destructive-operation gate ✅ Full
S12 The intelligence-freshness gate ✅ Full
S13 Shrink-policy: protecting enriched fenced content ✅ Core
S14 Bridge-refresh safety ✅ (policy) / ⚙ (operator Pre-Flight) Full
S15 The content scanner (agentteams.scan) ✅ Core
S16 Live-data redaction and feed sanitization ✅ Core
S17 The infrastructure-layers model ✅ (reference doc) Core
S18 Sandbox emission and privilege profiles ✅ (Linux bwrap deny-tested) / ⚙ (macOS Seatbelt UNVERIFIED, Windows design-only) Full
S19 The constitutional-gate hook ✅ (fail-closed under confined/exclusive; fail-open by default) Full
S20 Threat-intelligence watch ✅ Core
S21 Red-team methodology ✅ Core
S22 Integrity manifests ✅ Full
S23 Provenance stamps ⚙ (library / pattern, not auto-wired into the default emit path) Core
S24 Backups and baselines ✅ Full
S25 Defense-in-depth: how the layers compose ✅/⚙ Core
S26 Glossary ✅ Core
S27 Sources ✅ Light