Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/TrentM6/baseline-corenpx agentmods add skills/trentm6/baseline-core/skill-buildingWrote 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/trentm6/baseline-core/skill-building)<a href="https://agentmods.dev/skills/trentm6/baseline-core/skill-building"><img src="https://agentmods.dev/badge/skills/trentm6/baseline-core/skill-building/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/trentm6/baseline-core/skill-building"><img src="https://agentmods.dev/badge/skills/trentm6/baseline-core/skill-building.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 | $0.00033 | $0.04897 |
| Opus 5.5 | $0.00013 | $0.01959 |
| Sonnet 5.5 | $0.00007 | $0.00979 |
| Haiku 4.5 | $0.00003 | $0.00490 |
Grade A, and why
skill-building 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 yesterday.
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 — 436 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill Building
Skill building is the core methodology for AI Workflows engagements. It's the process of capturing expertise, codifying it into repeatable skills, and enabling client teams to use AI effectively for their specific work.
Core Principles
-
Curation is the architecture. A library's value is not the number of skills in it. Every skill added makes picking the right one harder, so the default answer to "should this be a skill?" is no until the evidence says otherwise.
-
Mine, don't import. Skills worth having come from your own corrections, rewrites, and repeated explanations. Someone else's library encodes their judgment about their work — the transferable part is the format, not the value.
-
Skills encode judgment, not just procedures. A skill shouldn't just say "do X, Y, Z." It should teach how to think about the domain — principles, trade-offs, when to escalate.
-
Portable across contexts. A good skill works with different business contexts. The skill stays the same; the context swaps (Baseline vs. client, client A vs. client B).
-
Lean core, deep references. SKILL.md should be 270-550 lines — the essential workflow. Deep content (frameworks, terminology, examples) lives in reference files that load on demand.
-
Progressive disclosure. Don't overwhelm. Core principles first, detailed methodology when needed, reference material for specific situations.
-
Quality checks are measurable. "Is it good?" isn't a quality check. "Does the main point land in 5 seconds?" is.
-
A written rule is the weakest lever. Prose in a skill file is a request the model can decline; a check that fails the build is not. For any rule the skill wants to hold, work down the Enforcement Ladder and write prose only when the rungs above it can't carry it.
-
Anti-patterns come from real mistakes. Don't invent theoretical failures. Document what actually goes wrong.
-
Skills must be maintainable. If the client can't update and extend the skill after handoff, it will decay. Design for maintenance.
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.
- yesterday First seen · 436 lines · 33 tokens per session scan A ca9e77f1c04d
skill-building is a skill published in the GitHub repository TrentM6/baseline-core (23 stars, last pushed 2d ago), licensed MIT. It adds 33 tokens to every session and 4,897 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 2026-11-17.
skill-building compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| skill-buildingTrentM6/baseline-core | Skill | 23 | 2d ago | MIT | A | 33 | Claude Code, Codex |
| hr-onboardingnexu-io/open-design | Skill | 108k | today | Apache-2.0 | A | 62 | Claude Code, Codex |
| book-mirrorgarrytan/gbrain | Skill | 33k | today | MIT | A | 138 | Claude Code |
| lean-formalizewanshuiyin/Auto-claude-code-research-in-sleep | Skill | 19k | today | MIT | A | 81 | Claude Code |
| exam-forecastanthropics/claude-for-legal | Skill | 10k | 2d ago | Apache-2.0 | A | 73 | Claude Code |
| eli5companion-inc/feynman | Skill | 10k | 4d ago | MIT | A | 63 | Claude Code, Codex |
| deck-course-modulenexu-io/html-anything | Skill | 10.0k | 3d ago | Apache-2.0 | A | 25 | Claude Code, Codex |