dsh-abyss
深海事务所 (Abyss) — an operations console for DeepSeek Harness agent fleets: every agent is a character at a desk, every line comes from the durable session log, and any past case replays and reports from disk
81 results
深海事务所 (Abyss) — an operations console for DeepSeek Harness agent fleets: every agent is a character at a desk, every line comes from the durable session log, and any past case replays and reports from disk
模型能力画像: 回顾式分析 ~/.dsh 历史会话 → 按模型的能力画像(擅长面/易翻车/建议用法) + 会话区只读观察窗 tab
Prompt-cache observability and logical-input diagnostics for DeepSeek Harness
Read-only Sentry issue and event tools for DeepSeek Harness.
A continuous, model-assisted semantic debugger for DeepSeek Harness trajectories.
Token usage dashboard, provider balances, and subscription quotas for the dsh web GUI
Read-only Grafana metrics and alert tools for DeepSeek Harness.
Privacy-conscious Prometheus metrics and a Grafana dashboard for DeepSeek Harness
Replay and share redacted DeepSeek Harness sessions as standalone HTML
Task Flow / Execution Graph plugin for DeepSeek Harness (DSH): visualize a single task's full execution — turns, LLM steps, tools, skills, subagents, retries, parallel branches — as an interactive DAG with live updates and trajectory linkage.
Local JSONL run traces and a live viewer for DeepSeek Harness sessions
DSH-Plugin: structural observability layer — real tool/skill usage recording (tools/result), skill-catalog audit, and a watchdog with health/error events. No model self-reporting.
Embedded yiTrace plugin for DeepSeek Harness turns, model steps, and tool calls
OpenTelemetry traces backend for the DeepSeek Harness telemetry seam: folds agent session events into OTLP span trees
Self-contained token usage statistics for DeepSeek Harness.
Design, observe, diagnose, and reuse evidence-backed multi-Agent work in DeepSeek Harness
Live Mission Control for DeepSeek Harness agents, tools, subagents, and token usage
A tiny behavioral mood ring for your AI coding agent: observes session behavior and shows a four-state mood status light in the session header
Read-only live subagent process inspector for DeepSeek Harness Web
X-ray for your DeepSeek Harness — diagnostics for what's actually loaded, why, and what it costs: per-plugin context-tax attribution, per-request token ledger, skill catalog pricing, dependency cascades.
DSH web plugin: a cross-session Window Stats view tab showing conversation progress and token usage for all sessions.
Explain failed DeepSeek Harness agent runs locally, then produce a redacted recovery plan. No model required.
dsh-codex-observability
A Jupyter-shaped, Git-backed research surface for DeepSeek Harness: session lineage, stable script attempts, visual evidence, provenance, safe previews, and explicitly confirmed reruns.