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 dimosgit/grounded-knowledge-engine --skill complex-task-workflowgit clone --depth 1 https://github.com/dimosgit/grounded-knowledge-engineWrote 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/dimosgit/grounded-knowledge-engine/complex-task-workflow)<a href="https://agentmods.dev/skills/dimosgit/grounded-knowledge-engine/complex-task-workflow"><img src="https://agentmods.dev/badge/skills/dimosgit/grounded-knowledge-engine/complex-task-workflow/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/dimosgit/grounded-knowledge-engine/complex-task-workflow"><img src="https://agentmods.dev/badge/skills/dimosgit/grounded-knowledge-engine/complex-task-workflow.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00047 | $0.00615 |
| Opus 5.5 | $0.00019 | $0.00246 |
| Sonnet 5.5 | $0.00009 | $0.00123 |
| Haiku 4.5 | $0.00005 | $0.00061 |
Grade A, and why
complex-task-workflow 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 9d 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 — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Token-First Complex Task Workflow
Follow AGENTS.md; user instructions remain authoritative.
Confirm the entry condition
The always-loaded repository contract owns the cheap routing decision. This skill is explicit-only because a mistaken automatic load costs an extra context-bearing turn. If it was selected for one clear path or a small synthetic task, stop here and return to ordinary execution: do not read references, inspect models, plan, or delegate.
Continue only for either:
- two substantial independent deliverables with disjoint writes, separate checks, and enough work for each to repay agent startup;
- an explicit user request to trade additional tokens for parallel elapsed time.
Invocation does not make delegation efficient by itself.
Choose the cheapest topology
Keep one coordinator by default. Search narrowly, batch independent tool calls, read only relevant files, run the smallest meaningful checks, and stop when they pass.
Without reliable usage telemetry, assume every fresh agent increases total tokens. In token-priority mode, use at most one worker and only when its separate context lets the coordinator avoid a large, distinct body of input. Never repeat the worker's exploration. In explicit time-priority mode, at most two workers may overlap when they have disjoint files, settled interfaces, and meaningful independent work. Never delegate recursively.
Route only a real worker
Do not build a role table. Immediately before the first worker call, read the active provider reference only if the native tool schema is insufficient: Codex or Claude Code.
Use native model and reasoning arguments, selecting the lowest-capability exposed option adequate for the deliverable and check. A model name in prose selects nothing. Unknown effective settings stay unknown. Same-model workers may save parallel time; they are not a token or cost optimization.
Use a stronger model only for one narrow unresolved decision when likely rework costs more than the consultation. Do not add a reviewer by default.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 9d ago First seen · 73 lines · 47 tokens per session scan A 82a04cfd890f
complex-task-workflow is a skill published in the GitHub repository dimosgit/grounded-knowledge-engine (5 stars, last pushed 10d ago), licensed MIT. It adds 47 tokens to every session and 615 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-10-11.
Other skills, from other repositories
github-issue-lifecycle
A GitHub Issue workflow for taking a code change from request to review, approval, testing, and closure. It can also keep a GitHub Project board—the task board attached to a repository—in sync.
capture-task
Capture a new task — create a draft or pending Task record from a rough idea or detailed spec. Use when asked to "add a task", "log a bug", "create a task", or "add to backlog".
refine-backlog
Refine the backlog — screen drafts, remove duplicates, fill in missing details, classify, link, and move to pending.
implement-task
Pick up and work on a binder Task — investigate bugs, discuss open questions, design solutions, write tests, or implement code. Use when asked to "work on", "pick up", "implement", or "fix" a task.
plan
Show the current work picture - the ready beads (agent tasks with no blockers) and the vault's open loops - to decide what to pick up next. Use at the start of a session or when planning the next chunk of work.
domain-status
Summarize the current state of one vault domain (CloudOps, DevOps, FinOps, SecOps, Architecture) - its notes, what changed recently, and the open loops that name it. Use before a weekly review, before picking up work in a domain, or when asked "where are we on X".