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 skills add echoVic/blade-code --skill goals-tasks-and-schedulinggit clone --depth 1 https://github.com/echoVic/blade-codeWrote 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/skills/echovic/blade-code/goals-tasks-and-scheduling)<a href="https://agentmods.dev/skills/echovic/blade-code/goals-tasks-and-scheduling"><img src="https://agentmods.dev/badge/skills/echovic/blade-code/goals-tasks-and-scheduling.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 5.1 | $0.00143 | $0.02790 |
| Opus 5 | $0.00072 | $0.01395 |
| Sonnet 5 | $0.00029 | $0.00558 |
| Haiku 4.5 | $0.00014 | $0.00279 |
Grade A, and why
knowledge-agent-execution-and-orchestration-goals-tasks-and-scheduling 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 5d 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Module Structure
该节点包含三类不同的“持续工作”:每 Session 一个长期 Goal、供 Agent/团队协调的轻量任务列表,以及由 Web/CLI/调度器创建并独立运行的顶层任务 Session。
Directory Layout
packages/cli/src/goals/— Goal sidecar、状态机、提前停止检测和 continuation promptpackages/cli/src/agent/runtime/TaskScheduler.ts—blade serve拥有的定时扫描与 dispatchpackages/cli/src/agent/runtime/scheduleTiming.ts— cron、interval、once 解析与下次执行计算packages/cli/src/services/ScheduleStore.ts— 全局 schedule JSON 存储packages/cli/src/services/SessionTaskService.ts— 顶层任务 Session 与 worktree 创建packages/cli/src/tools/builtin/goal/— Goal 查询、创建和完成候选工具packages/cli/src/tools/builtin/task/— Agent 任务列表与子代理 Task 入口packages/cli/src/commands/schedule.ts— CLI schedule 管理与远程手动触发packages/cli/src/server/routes/task.ts— 顶层任务创建、更新、重试、diff 与交付 APIpackages/cli/src/server/routes/schedule.ts— schedule CRUD 与手动运行 APIpackages/cli/web/src/components/tasks/,packages/cli/web/src/components/kanban/,packages/cli/web/src/components/schedules/— Web 任务与调度表面
Key Entry Points
GoalStore.create()/GoalStore.requestCompletion()inpackages/cli/src/goals/GoalStore.ts— Goal 启动与完成候选GoalStore.finalizeVerifiedCompletion()inpackages/cli/src/goals/GoalStore.ts— 宿主授权的最终完成TaskListManager.mutate()inpackages/cli/src/tools/builtin/task/TaskListManager.ts— 任务列表串行读改写SessionTaskService.createSessionTask()inpackages/cli/src/services/SessionTaskService.ts— 顶层任务和隔离 workspace 创建TaskScheduler.tick()/TaskScheduler.fire()inpackages/cli/src/agent/runtime/TaskScheduler.ts— 定时触发与账本推进computeNextRun()inpackages/cli/src/agent/runtime/scheduleTiming.ts— 三类 trigger 的统一时间计算
Gotchas
UpdateGoal({status:"complete"})只把状态改为verifying并增加 attempt;必须存在 fresh verifier Session ID 和 SHA-256 evidence digest 的 PASS,宿主才允许complete(packages/cli/src/tools/builtin/goal/goalTools.ts,packages/cli/src/goals/GoalStore.ts,git:0eff1b2a)- reserved verifier 的启动本身证明宿主已持久接受 completion candidate;审计期间的
verifying/pending是等待该 verifier verdict 的必要中间态,不能被解释成执行 Agent 未调用UpdateGoal complete,但也不能替代对实际交付物的验证 (packages/cli/src/agent/loop/goalCompletionVerification.ts,packages/cli/src/agent/subagents/builtinGoalVerificationAgent.ts) - Goal 已有 PASS 后发生 workspace mutation、用户 steering 或 Stop hook continuation 会使证据失效;进程重启也不信任没有 exact finalization receipt 的旧 PASS (
packages/cli/src/agent/loop/executeLoopGenerator.ts,docs/reference/goal-completion-verification.md) - Goal 最终 assistant commit 与 sidecar 完成不是一个文件事务;恢复只在 goal ID、attempt、verifier ID、evidence digest 和 goal revision 全匹配 receipt 时幂等补写 complete (
packages/cli/src/goals/GoalStore.ts,docs/reference/goal-completion-verification.md,git:ebd505c9) - 同一脱敏 verifier 反馈连续 3 次会把 Goal 置为 blocked;不同反馈重置计数,PASS、编辑或显式 resume 清除 stall 状态 (
packages/cli/src/goals/GoalStore.ts,packages/cli/src/goals/types.ts,git:490df8c5) - premature-stop 只检查最终非空段落中一组保守的英文起始模式;同一模式连续 3 次才自动 blocked,模式改变会从 1 重新计数 (
packages/cli/src/goals/prematureStop.ts,packages/cli/src/goals/GoalStore.ts) - Goal token 使用按每次 continuation 的 run metadata 累加,达到显式预算立即进入
budget_limited;结束一个模型回合不会隐式暂停或完成 Goal (packages/cli/src/agent/Agent.ts,packages/cli/src/goals/GoalStore.ts) - 通用 TaskList 的状态流转主要由工具提示约束,存储层允许直接设置任一合法状态;调用方若绕过
pending → in_progress → completed,不会由TaskListManager自动拒绝 (packages/cli/src/tools/builtin/task/taskListTools.ts,packages/cli/src/tools/builtin/task/TaskListManager.ts) - 通用 TaskList 可添加不存在的
blockedByID,结果会永久不可认领;Team 创建路径额外要求依赖只能引用更早任务,不能假设这个校验存在于底层 manager (packages/cli/src/tools/builtin/task/TaskListManager.ts,packages/cli/src/agent/teams/TeamRuntime.ts) - Schedule 只有
blade serve内的TaskScheduler会自动触发;CLI 只读写共享文件,blade schedule run也必须调用正在运行的 server (packages/cli/src/commands/schedule.ts,packages/cli/src/agent/runtime/TaskScheduler.ts) - 同一 schedule 的 overlap 防护和 tick 重入锁都只在当前 server 进程内;
ScheduleStore的写链也只保证进程内串行,多个进程同时管理 schedule 仍是 last-writer-wins (packages/cli/src/agent/runtime/TaskScheduler.ts,packages/cli/src/services/ScheduleStore.ts) - dispatch 抛错仍会推进 recurring 的下一次时间,once 则消费唯一机会并停用;错误路径不增加
runCount,因此该字段统计成功 dispatch 而不是所有尝试 (packages/cli/src/agent/runtime/TaskScheduler.ts) - 手动运行 recurring schedule 保留原
nextRunAt,不会把 cadence 从手动触发时刻重新锚定;手动运行 once 仍会将其停用 (packages/cli/src/agent/runtime/TaskScheduler.ts) ScheduleStore对缺失或损坏文件都返回空列表并记录 warning;后续写操作可能以空集合覆盖损坏内容,诊断数据丢失前应先检查日志与原文件 (packages/cli/src/services/ScheduleStore.ts)parseIntervalMs()接受秒级输入,但持久 schedule 的validateTrigger()要求至少 60 秒;只调用解析器通过不代表 trigger 可创建 (packages/cli/src/agent/runtime/scheduleTiming.ts)
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.
- 5d ago First seen · 77 lines · 143 tokens per session scan A 2e58a8a627fc
knowledge-agent-execution-and-orchestration-goals-tasks-and-scheduling is a skill published in the GitHub repository echoVic/blade-code (178 stars, last pushed 2d ago), licensed MIT. It adds 143 tokens to every session and 2,790 once invoked, about $0.0007 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 2026-09-03.
Other skills, from other repositories
linearis
Manage Linear.app work from the command line with the linearis CLI (bins linearis / linear), which outputs JSON: issues/tickets, projects, cycles (sprints), milestones, initiatives (roadmap), documents, labels, teams, users, and issue discussions/comments. Use when the user mentions Linear, a ticket identifier like…
autogpt-agents
Autonomous AI agent platform for building and deploying continuous agents. Use when creating visual workflow agents, deploying persistent autonomous agents, or building complex multi-step AI automation systems.
linear
Linear: manage issues, projects, teams via GraphQL + curl.
github-issues
Create, triage, label, assign GitHub issues via gh or REST.
stale-sweep
Sweep the googleapis/mcp-toolbox repo for issues and PRs with no real activity in N days (default 60), sort each by whose silence it is (the author's, ours, or nobody's), and draft the nudge or close comment. Use whenever a maintainer asks for a stale sweep, backlog cleanup, or an SLO check, e.g. "stale sweep", "find…
clinical-reports
Write comprehensive clinical reports including case reports (CARE guidelines), diagnostic reports (radiology/pathology/lab), clinical trial reports (ICH-E3, SAE, CSR), and patient documentation (SOAP, H&P, discharge summaries). Full support with templates, regulatory compliance (HIPAA, FDA, ICH-GCP), and validation…