Skip to main content
DE

Deliberate Ensemble

Research Archive

◈Home⊕Projects & Builds🎙️MeshCast▶Videos☰Library◉System Pulse◇Nexus Graph⋇Papers⚖Governance⊛About→Start Here📊Observability▤Logs◆Agents
Documents16,501
Tags123
Categories157
← Back to Library
doc·docs

Defensive patterns

View on GitHub →
Verification
Path: docs/defensive-patterns.mdModified: 9/2/2026

Location

self-organizing-library/docs

Content

Loading content...

Cross-References

Agent Note: Plan mode — a logged per-agent session modelinkAgent Note: plan mode——记录到日志的逐 agent 会话模式linkAgent Note: stdin + extra env on the bash seamlinkAgent Note: Subagent provider-lifecycle events — subagent/provider-added / subagent/provider-removedlinkAgent Note: Cooperative tool cancellation at the registry boundarylinkAgent Note: Code Mode — the model writes TypeScript against the tool registrylinkAgent Note: Code Mode — the model writes TypeScript against the tool registrylinkAgent Note: Drop bash full-output spill fileslinkReviewing a DeepSeek-Harness PRlinkReviewing a DeepSeek-Harness PRlinkFinding DeepSeek Harness SimplificationslinkAGENTS.mdlink防御性模式linkTesting policylink防御性模式linkCode Runtimelink@deepseek-ai/dsh-hook-protocollink
d3b0fef4b1b7a1f5

Related by Tag

Architecture DocumentationTechnical Brief: AI Ensemble Intelligence SystemTheory: AI Ensemble Intelligencefeature:pull request template

Metadata

Extension.md
Size3.6KB
Repodeepseek-harness

Document ID

d3b0fef4b1b7a1f5