Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/ccdawn/vibelution/conversation-flow-mapgit clone --depth 1 https://github.com/CCDawn/VibelutionWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/ccdawn/vibelution/conversation-flow-map)<a href="https://agentmods.dev/agents/ccdawn/vibelution/conversation-flow-map"><img src="https://agentmods.dev/badge/agents/ccdawn/vibelution/conversation-flow-map.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.04082 |
| Opus 5 | $0.00000 | $0.02041 |
| Sonnet 5 | $0.00000 | $0.00816 |
| Haiku 4.5 | $0.00000 | $0.00408 |
Grade A, and why
conversation-flow-map scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
对话链路地图
这份文档用于维护 Chat/Coding 从用户发送消息到模型内容返回 UI 的主链路。它不是重构方案，而是后续学习、诊断和小步优化时的共同地图。
范围
主路径：
ChatCodingRoute 乐观提交 -> FastAPI sessions route -> session_service 提交/调度/worker（实现 claim 见 core/web/services/session/*） -> agent.py 单轮执行 -> core.llm invoke/stream -> UI capture -> turn journal/live output -> SSE -> 前端 active-turn layer -> ConversationView。
结构拆分历史说明（非现行规范）：docs/archive/plans/2026-06-07/2026-07-21-backend-structure-p0-completion.md。Team 工作流产品面 claim 见 core/web/services/team_workflow/README.md（routes 经 team_workflows/ 包 + team_workflow_orchestration_service facade）。
不覆盖：
- Chat room 多人轮次。
- Self-evolution 编排本身；只有它复用 hidden/direct session 时才回到本链路。
- provider adapter 内部细节；这里仅追到共享
core.llm.invocation入口。
稳定入口
| 层 | 入口 | 职责 |
|---|---|---|
| 前端提交 | web/src/routes/ChatCodingRoute.tsx |
发送 POST /api/sessions/{sessionId}/messages，带 Prefer: respond-async；先乐观更新 UI，再通过 stream/cache 校准。 |
| API route | core/web/routes/sessions.py |
拥有 /api/sessions/{id}/messages 和 /api/sessions/{id}/events。 |
| 提交 owner | core/web/services/session/submit.py::submit_session_message（facade re-export：session_service；见 session/README.md） |
校验输入，解析附件/引用，开启 turn，写初始 journal，发布首个 snapshot，并调度后台执行。 |
| 调度 owner | core/web/services/session/schedule.py::_schedule_session_turn（facade re-export） |
会话/Agent 并发队列、executor 提交与释放。 |
| worker owner | core/web/services/session/worker.py::_run_session_turn（facade re-export） |
解析 runtime/model/context，创建 chat agent，捕获 UI stream，运行 agent turn。 |
| capture owner | core/web/services/session/stream_capture.py::_capture_session_ui_stream（facade re-export） |
UI thought/response/tool 批处理写入 live_output 与 journal 片段。 |
| persist owner | core/web/services/session/persist.py::_persist_session_turn_result（facade re-export） |
持久化最终 assistant / turn_*，清理 live output，发布终态 detail。 |
| Agent turn | agent.py::run_single_turn |
执行一轮 agent thought/action loop，返回结构化 result dict。 |
| LLM 调用 | agent.py::_invoke_llm + core/llm/invocation.py |
统一经过 streaming/invoke helper，并携带 invocation metadata 与 prompt-cache partition。 |
| 事实源 | core/chat/turn_journal.py |
append-only turn 事件日志，用于 replay、模型上下文和可见消息投影。 |
| 前端流式层 | web/src/routes/sessionAssistantDeltaScheduler.ts 和 web/src/routes/chatActiveTurnLayer.ts |
平滑消费 assistant_delta，在最终 session_detail 到来前维护 live assistant 响应。 |
| 最终渲染 | web/src/components/conversation/ConversationView.tsx |
主路径 package_cells（turnItems → codexTranscript.cells）；无包时 legacy 走 content/timeline。 |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 132 lines · 0 tokens per session scan A 4036e962bf3b
conversation-flow-map is an agent published in the GitHub repository CCDawn/Vibelution (21 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 4,082 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
memory-consolidator
Use this agent for background memory consolidation. Typical triggers include asyncRewake pipeline signaling that N turns have accumulated since last consolidation, and explicit dispatch after memory-seed completes. See "When to invoke" in the agent body for worked scenarios.
Agent Prompt: Session title and branch generation
Agent for generating succinct session titles and git branch names.
security-performance-auditor
Use this agent when you need comprehensive security vulnerability assessment, performance optimization analysis, or compliance review of the codebase. Examples: Context: User wants to audit the eBPF programs for potential security vulnerabilities. user: 'Can you check our eBPF programs for any security issues?'…
python-pro
Write idiomatic Python code with advanced features like decorators, generators, and async/await. Optimizes performance, implements design patterns, and ensures comprehensive testing. Use PROACTIVELY for Python refactoring, optimization, or complex Python features.
data-scientist
Data analysis expert for SQL queries, BigQuery operations, and data insights. Use proactively for data analysis tasks and queries.
opencode
Point OpenCode at a local rapid-mlx server. OpenCode is a Claude-Code-like terminal coding agent that speaks the OpenAI-compatible chat completions API (POST /v1/chat/completions) via the @ai-sdk/openai-compatible provider.