dsh-memory-gate
Retrieved ≠ injected: CBDC-gated memory for DeepSeek Harness — decides how memory is USED (use/verify/ignore decisions, feedback learning, full audit); local SQLite + FTS5, bounded, no extra model call
207 results
Retrieved ≠ injected: CBDC-gated memory for DeepSeek Harness — decides how memory is USED (use/verify/ignore decisions, feedback learning, full audit); local SQLite + FTS5, bounded, no extra model call
User-set session/global must-do and must-not requirements injected at conversation start (Baize).
DeepSeek Harness plugin (temporary, pre-upstream-fix): automatically injects a CPython tempfile shim (PYTHONPATH -> sitecustomize) into every confined shell command on the Windows sandbox, so python/pytest tempfile use works with zero extra tools, zero model-context overhead, and zero escalation.
DSH v0.18.x LAN access plugin: binds the web server to 0.0.0.0 (bypassing the official --host 0.0.0.0 CLI guard at the supported composition layer) and injects a browser shim that fixes the crypto.randomUUID() secure-context crash which breaks the client RPC layer on plain-HTTP non-loopback origins (phone browsers).
Phase 1 Claw identity: inject core workspace files into the agent prompt. 把核心人设文件打进 Claw Agent 提示词。
Post-compaction premise-drift guard plugin for DeepSeek Harness: after a compaction summary drops a critical literal anchor (path, value, error code) it injects a one-shot notice telling the model what it may have lost and how to recover it from the log
dsh-experience-flywheel · DSH 经验飞轮插件 — agent cross-session experience memory + dual-acceptance gates (auto query/inject via agent/pre-step middleware; exit-1 gate scripts)
Auto-resumes dsh sessions whose turn was left mid-flight by a web restart or a transient upstream failure (rate limits, 5xx, network faults, quota/billing errors) by injecting a 'continue' message, with a dead-loop guard that hands control back if the retry produces nothing.
DSH LLM response-stream injection/pollution filter: hard-block rare Unicode scripts (Track A) and score-based disposal of control chars / protocol markers / script mixing / spam keywords (Track B) on the llm/stream waterfall.
Outbound LLM request header modification for DeepSeek Harness (DSH): client-identity header presets (Codex / Claude Code / opencode), per-request dynamic headers, X-Stainless fingerprint stripping, URL rewriting and Anthropic body patching at the fetch transport layer.
A safe external-skill discovery and adoption bundle for DeepSeek Harness
Inject a configurable dsh-ui status card instruction before every DSH agent reply
Unofficial community plugin: per-session Engram memory for DeepSeek Harness — spawns an engram MCP child per agent session rooted at the session workspace, registers mem_* tools per agent scope, and injects the Memory Protocol as a system-prompt section.
Runtime security guard for DeepSeek Harness (DSH) — blocks command injection, SSRF, credential exfiltration, and destructive operations at runtime.
Task-to-skill pairing with a laziness ladder and a safe install cellar. Never overpours, never serves untasted. DSH · Claude Code · Codex
Automatic TMCRA recall, injection, and role-separated writeback for DeepSeek Harness
An integrated starter plugin for newly installed DeepSeek Harness (DSH). Combines whale-girl persona injection, desktop pet, balance widget, notes, reminders, shared long-term memory, and low-token memory tools.
Enforced pain-point-check guard plugin for DeepSeek Harness: after two non-converged experiments on one problem it injects the three questions, denies non-investigative tool calls until they are answered in the reply text, and blocks same-direction retries
DSH web extension: live subagent monitor + role-based subagent routing + Claude Code style .dsh/agents/*.md persona injection
Claude Code-style `paths:` rule injection for DeepSeek Harness (DSH): inject rules from ~/.dsh/rules, <project>/.dsh/rules and <project>/.claude/rules into the model context.
Portable DSH WebUI skill security inspection with local rules and optional model review
Focus board for DeepSeek Harness agents: durable, model-maintained notes in the session workspace that pin the objective, constraints, and decisions across compaction and sessions — with structured entry tags (kind/status), a grouped summary view, automatic context injection, an archive on clear, and an optional read-only web panel
A runtime prompt-injection shield for DeepSeek Harness tool results.
Ponytail for DSH — the lazy senior dev. Live intensity switching (off/lite/full/ultra) with a toggle in the composer, injecting a minimal-code ruleset before every step to cut token usage.