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 bestdeejay-design/agent-skills --skill skill-forgegit clone --depth 1 https://github.com/bestdeejay-design/agent-skillsWrote 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/bestdeejay-design/agent-skills/skill-forge)<a href="https://agentmods.dev/skills/bestdeejay-design/agent-skills/skill-forge"><img src="https://agentmods.dev/badge/skills/bestdeejay-design/agent-skills/skill-forge.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 5.1 | $0.00173 | $0.02348 |
| Opus 5 | $0.00086 | $0.01174 |
| Sonnet 5 | $0.00035 | $0.00470 |
| Haiku 4.5 | $0.00017 | $0.00235 |
Grade A, and why
skill-forge 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 7d 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 — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.
skill-forge
A compass, not a dictator. This skill helps you design skills that are maximally technological (engineered, reliable, composable), creative (generative, varied, distinctive), and aesthetically disciplined (intentional, not templated). It separates the obligatory base — rules you must not break — from accents — amplifiers you choose per task.
Russian triggers:
создай скил,новый скил,улучши скил,сделай скил креативнее,эстетичный скил,технологичный скил,скилл шаблонный,перепиши SKILL.md,мета-скилл,проверь скилл перед релизом.
Level 0 — Hard rules (the base, never break these)
These come from the official Agent Skills specification (agentskills.io) plus our own working gates. They are non-negotiable.
- Frontmatter:
name(≤64 chars,a-z 0-9 -, no double hyphens, MUST match the folder name) anddescription(≤1024 chars, MUST state what the skill does AND when to use it). Put all "when to use" info indescription, never in the body. Make triggers a little "pushy" to fight undertriggering — list synonyms and casual phrasings. - Progressive disclosure: keep
SKILL.mdunder 500 lines. Bundle code inscripts/(deterministic, repetitive tasks), docs inreferences/(loaded as needed; add a TOC if >300 lines), outputs inassets/. Reference files one level deep only. - Voice: write in the imperative. Explain the why instead of heaped MUST/ ALWAYS/NEVER — all-caps rigidity is a yellow flag. The model is smart; reason with it.
- Principle of Lack of Surprise: no malware, no deceptive or malicious content. A skill's intent must match its description.
- Single data contract: one spec file (e.g.
deck.json) is the source of truth; every stage reads/writes it so artifacts are regenerable. - Mandatory gates: nothing is "done" until automated checks PASS. At minimum: structural/format validation, content QA, and — for any visual output — a real-browser or render gate. Require EVIDENCE, not assertion: the skill must show the command it ran, its output, and the exit status. Use SCOPED CLAIMS — separate failures the change caused from pre-existing ones ("3 pre-existing failures, my change introduced none"). For unattended runs, back the gate with a deterministic Stop hook or a /goal evaluator so a turn cannot end on a false "done".
- Reproducibility: deterministic seed so the same input yields the same output on another day.
- Anti-template baseline (our enforcement, mechanical): ≤2 typefaces, ≤4
accent roles (keep 60-30-10), WCAG contrast (text ≥4.5:1, large/non-text
≥3:1), zero raw hex outside
:root(every color is a token), one mood per artifact, one idea per slide, assertion headlines (full sentence with a verb), words-per-slide budget.
What ships with it
2 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.
- 7d ago First seen · 152 lines · 173 tokens per session scan A 2332addb0844
skill-forge is a skill published in the GitHub repository bestdeejay-design/agent-skills (5 stars, last pushed 3d ago), licensed MIT. It adds 173 tokens to every session and 2,348 once invoked, about $0.0009 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 2026-08-31.
Other skills, from other repositories
workflow
Use when a task is too large for turn-by-turn orchestration and should run through the big-task workflow lane: system-wide changes, large migrations, repo-wide audits, high-confidence verification, or tasks explicitly asking to run a workflow. Claude Code uses native dynamic workflows; Codex, OpenCode, and Grok use…
skill-compiler
Automatic solved-to-skill compiler — detects novel task completions and autonomously drafts new SKILL.md files. Stolen from Hermes Agent's learning loop (NousResearch, 2026-05-11).
context-compactor
9-section context compression with analysis scratchpad. Adapted from Claude Code's /compact system (2026-03-31).
daemon-loop
Autonomous recurring agent tasks — converts workflows into persistent background daemons that run on intervals. Stolen from Boris Cherny's Claude Code /loop pattern (2026-03-31).
trade-journal-analyzer
Unified post-trade analytics: journal pattern extraction + drawdown classification. Absorbs: drawdown-classifier.
Deep Research Loop
Multi-step web research, compilation, and synthesis workflow. Scrapes multiple sources, cross-references claims, and produces a structured research brief.