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 skills/kachofugetsu09/akashic-agent/plugin-systemnpx skills add kachofugetsu09/akashic-agent --skill plugin-systemgit clone --depth 1 https://github.com/kachofugetsu09/akashic-agentWhat 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 | $0.00036 | $0.01559 |
| Opus 5 | $0.00018 | $0.00779 |
| Sonnet 5 | $0.00007 | $0.00312 |
| Haiku 4.5 | $0.00004 | $0.00156 |
Grade A, and why
plugin-system 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 2d 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 — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Akashic 插件系统
优先完成明确的插件请求。创建或改写 source、加入 Skill/MCP、安装候选或递归验证时，先加载 develop-akashic-plugin；这里只负责安装链、状态边界和 turn 结果。
事实来源与边界
┌─ ~/.akashic-plugin/manifest.toml
│ └─ 全局安装清单与启用状态
├─ ~/.akashic-plugin/cache/<marketplace>/<plugin>/.artifacts/
│ └─ 不可变 installed artifact 与 generation pointer
├─ <workspace>/plugin-data/<plugin>-<marketplace>/
│ └─ 插件配置与持久状态
└─ <workspace>/runtime/plugin-reloads.sqlite3
└─ reload、candidate、恢复与 turn rollout 证据
外部 source 的 canonical 根必须包含 akashic.plugin.toml 和其 entrypoint 指向的 v3 module。manifest 的 schema_version=1、api_version=3、name、version、entrypoint 在 import 前校验；module 还必须导出同值的 api_version/name/version 与精确 apply(ctx, config)。Skill/MCP 通过 source 的静态 root 和 typed service 进入 generation catalog。
不要查找或创建 registry.json、.aka-plugin/plugin.json、manifest.yaml、插件级 mcp/servers.json 或 workspace 手工 Skill owner。不要直接编辑 cache、pointer、全局 manifest、workspace Skill 软链接或正式 plugin-data。
主动信息源也只组合普通服务：TIMERS 驱动来源私有 poll，离散事实提交 CONTENT，当前状态保留在插件私有 cache，候选行动进入 DRIFT，完整推理由 BACKGROUND_JOBS 创建普通 Turn。不要创建 proactive catalog、私有 lifecycle family 或 MCP 聚合桥。
Agent 可用动作
plugin-install 安装或更新本 turn 的候选
plugin-uninstall 登记本 turn 结束后的卸载
plugin-revert 撤销本 turn 最近一次尚未提交的操作
不要手工编辑 manifest/cache/pointer，不要手工切换 generation、重启 Gateway 或用第二个进程绕过 turn 边界。stable、candidate、lease、排空、提交、恢复、能力投影和服务切换由 Core 拥有。
安装与更新
只从已提交 Git HEAD 安装；远程 source 必须先 push 对应 commit：
python main.py plugin-install \
--source <repo_or_url> \
--marketplace <marketplace>
正常链路：
source test → commit/push → plugin-install
→ attached child 的真实行为 oracle
├─ pass → 正常结束父 turn → Core 自动提交 → 下一 turn 生效
└─ fail → plugin-revert → 修复 source 后递归
安装成功只表示候选准备并绑定当前父 turn；父 turn 仍使用旧 stable，attached child 自动使用候选。子 turn 不指定 runtime、不 detach、不直接选择 candidate。结果必须保存 candidate identity、reload transaction、child identity、tool trace 和 terminal；只看命令返回或 catalog 不足以证明功能。
Skill、MCP 与服务检查
┌─ Skill
│ └─ source root、SKILL.md、references、catalog source、真实触发轨迹
├─ MCP
│ └─ manifest command、requirements、required tools、candidate read-only tools、endpoint env
├─ managed service
│ └─ process identity、port_env、readiness、退出与隔离 plugin-data
├─ Channel
│ └─ descriptor、credential paths、stop/start ownership 与恢复
└─ rollout
└─ child terminal、tool items、reload journal、turn 后 generation
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.
- 2d ago First seen · 106 lines · 36 tokens per session scan A bfe1a97a95b5
plugin-system is a skill published in the GitHub repository kachofugetsu09/akashic-agent (486 stars, last pushed 2d ago), licensed MIT. It adds 36 tokens to every session and 1,559 once invoked, about $0.0002 per session on Opus 5. 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 skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
agent-host-chat-contributions
Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.