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 KaydenClark/LLM_Workbench --skill implement-specgit clone --depth 1 https://github.com/KaydenClark/LLM_WorkbenchWrote 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/kaydenclark/llm_workbench/implement-spec)<a href="https://agentmods.dev/skills/kaydenclark/llm_workbench/implement-spec"><img src="https://agentmods.dev/badge/skills/kaydenclark/llm_workbench/implement-spec/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/kaydenclark/llm_workbench/implement-spec"><img src="https://agentmods.dev/badge/skills/kaydenclark/llm_workbench/implement-spec.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 | $0.00020 | $0.00714 |
| Opus 5.5 | $0.00008 | $0.00286 |
| Sonnet 5.5 | $0.00004 | $0.00143 |
| Haiku 4.5 | $0.00002 | $0.00071 |
Grade A, and why
implement-spec 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 11d 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.
What it actually says
Start with an authorized Spec already sliced into Tasks. Resolve their records through the project manifest. Follow the Workbench Contract and its existing claim, verification and lifecycle procedures.
The goal is the entire Spec implemented on one assembly branch, with a PR into the project's integration branch marked ready for review.
The Dispatcher orchestrates; Workers implement every change and correction. Directors coordinate multiple Dispatchers and parallel Specs.
Task dependencies form a graph and determine which Tasks are ready. Run compatible Workers concurrently, respecting active claims and shared writers.
Communicate through context pointers to the Spec, Tasks, research notes and previous commits.
Steps
-
Read the Spec and Tasks to recover progress and understand dependencies. Directly connected Tasks from other Specs may unblock this Spec. Without a Director, delegate only those necessary Tasks to Workers, taking the smallest number of reversible actions and preserving original Spec links.
-
Optionally use an exploration subagent for source or documentation research. Save its notes in local, untracked context accessible to future Workers.
-
Create or resume the assembly branch from current project integration. Open a draft PR into integration after the first merge, linked to the Spec and included Tasks.
-
Dispatch implementer Workers, each in its own worktree and branch based on assembly. Each uses red/green TDD and required Task checks, merges the latest assembly tip before hand-back, rechecks affected work, and returns exact commits, proof, documentation and remaining gaps.
-
Use a merger subagent to merge verified results into assembly one at a time and check the resulting assembly SHA before releasing dependent Tasks. Delegate implementation fixes to Workers. Necessary cross-Spec results merge here too; their records and proof stay under their original Specs, which remain unfinished until their own completion requirements are met.
-
Recompute readiness from verified results contained in this assembly, without waiting for PR or tracker closure. A dependency is satisfied for this run only when its verified result is contained here. Start newly ready, compatible Workers. One state writer maintains Task proof and owning records.
-
Once all Tasks are assembled, arrange one full code-review and send fixes to one implementer Worker for one correction pass. Merge corrections through step 5, then run focused verification and required checks on the final assembly candidate. This is building-side evaluation; integration approval follows later. Unresolved findings keep the PR draft with a clear continuation point.
-
When evaluation and checks pass, mark the PR ready for integration review -> verify. Record the PR, final candidate, verification and next gate in the owning Spec and regenerate its projection so later Workbench activity can discover readiness.
-
Clean up implementer worktrees through the host's supported operation after preserving proof and verifying their commits are contained in assembly. Keep assembly available for review. Report the assembly branch and PR; the run ends before integration merge.
If required capabilities are missing or no safe work can advance, preserve state and report the blocker. The Dispatcher does not implement instead.
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.
- 11d ago First seen · 72 lines · 20 tokens per session scan A 9753f6671a89
implement-spec is a skill published in the GitHub repository KaydenClark/LLM_Workbench (2 stars, last pushed 2d ago), licensed MIT. It adds 20 tokens to every session and 714 once invoked, about $0.0001 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 2026-10-08.
Other skills, from other repositories
mcore-split-pr
Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.
shiplog
Recap of everything shipped since the last run - cross-repo PRs, security fixes, star deltas, and X traction, synthesized into a digest article and a ready-to-post shiplog in your voice.
connect
Choose which wiki this session uses, the team project's pages on the Nimbalyst server or the project's local wiki of files, and connect a repository to its team project. Use at the start of a task before any wiki tool, when a wiki tool returns reponotbound, ambiguousproject, pinmismatch, projectnotaccessible…
llxprt-issue-workflow
Use this skill when asked to address, fix, or work on a GitHub issue in the llxprt-code repository. Covers the complete issue lifecycle - branch setup, gh CLI usage, test-first planning, subagent delegation and review loops, the full verification cycle (including the zai-glm-flash smoke test), open code review (ocr)…
cuga-github-issues
Create GitHub issues for cuga-agent (bugs, features, epics, designs, and related work) against origin using gh, with epic → feature → issue hierarchy and GraphQL sub-issue linking. Use when the user says create new issues, report a bug, file a bug, feature request, new feature, create an epic, GitHub issue, sub-issue…
implement-issue
Implement a Linear brief in a named GitHub repo. Use when the brief skill is implement-issue or the user wants a new branch and draft PR.