PSA is a multi-classifier behavioral analysis engine for LLM responses. It classifies every sentence into behavioral postures, then derives metrics from the posture sequences to detect adversarial stress, sycophancy, hallucination risk, persuasion techniques, and input pressure — in real time.
Multi-agent collaboration suite for DeepSeek Harness: specialist roster with on-demand dispatch, roundtable, model comparison and a multimodal vision bridge — models via the official provider flow.
Behavior discipline for coding agents: read-before-write enforcement, failure convergence, write verification with auto-rollback, and a high-risk command gate. Ships FULL plugin/hook adapters for DSH, Claude Code, Codex, and Pi, plus a host-agnostic SOFT skill with ADAPTIVE adapter guidance.
Unofficial community fork of the Aqua glassmorphism theme for the DeepSeek Harness web UI (upstream: WYH66666666/DSH-Transparent-UI-Plugin, MIT). Adds DSH rc.7 compatibility. Not affiliated with DeepSeek.
Native DeepSeek Harness plugin: gates pwsh tool calls against PowerShell pitfalls via the official tools/pre-execute interception point, bundles the powershell-check skill
Topic memory for LLM agents — edited, not accumulated: a topic keeps the starting question, conclusion, impact and dependencies; process is not memory. OKF bundle for dsh, local-first, git-traceable, budgeted LLM-free injection.
The open Company State Runtime — version control for your project's state. Goals, decisions, work, and evidence outlive every chat, context window, and agent session. Ships the self CLI.
Game-style floating HUD for DeepSeek Harness: balance HP bar, context MP bar, official peak/valley pricing with countdown, auto-compaction, memory-carrying new conversation. ?????? HUD ??
Verdict-based context compaction for DeepSeek Harness — replaces lossy LLM summaries with fast keep/truncate/drop decisions from jev-latest; everything kept stays verbatim. Port of tamaratran/fast-jev-compaction.
