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 drafthq/draft --skill new-trackgit clone --depth 1 https://github.com/drafthq/draftWrote 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/drafthq/draft/new-track)<a href="https://agentmods.dev/skills/drafthq/draft/new-track"><img src="https://agentmods.dev/badge/skills/drafthq/draft/new-track/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/drafthq/draft/new-track"><img src="https://agentmods.dev/badge/skills/drafthq/draft/new-track.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Prompt Injection · line 228 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- high Tool Misuse · line 817 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 818 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
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 | $0.00078 | $0.06719 |
| Opus 5.5 | $0.00031 | $0.02688 |
| Sonnet 5.5 | $0.00016 | $0.01344 |
| Haiku 4.5 | $0.00008 | $0.00672 |
Grade A, and why
new-track scanned grade A with 1 finding 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 1mo 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
3. Use `curl`/`wget` to fetch any URLs mentioned (dashboards, error pages, API responses) How it starts
The opening of the file, as written. The whole thing — 897 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Create New Track
Create a new track (feature, bug fix, or refactor) for Context-Driven Development. This is a collaborative process — be an active participant providing guidance, fact-checking, and expertise grounded in vetted sources.
Feature Description: $ARGUMENTS
Red Flags - STOP if you're
- Creating a track without reading existing Draft context (product.md, tech-stack.md, .ai-context.md)
- Asking questions without contributing expertise or trade-off analysis
- Rushing through intake without probing deeper with "why"
- Generating spec/plan without user confirmation at checkpoints
- Skipping risk identification
- Not citing sources when giving architectural advice
Collaborative understanding, not speed.
Pre-Check
- Verify Draft is initialized:
ls draft/product.md draft/tech-stack.md draft/workflow.md draft/tracks.md 2>/dev/null
If missing, tell user: "Project not initialized. Run /draft:init first."
-
Check for
--quickflag in$ARGUMENTS:- If present: strip
--quickfrom$ARGUMENTSnow (before Step 1) and store the cleaned text as the working description for all subsequent steps. Proceed to Step 1, then go directly to Step 1.5: Quick Mode. - Quick mode is for: hotfixes, tiny isolated changes, work scoped to 1-3 hours
- If present: strip
-
Load full project context (these documents ARE the big picture — every track must be grounded in them):
- Read
draft/product.md— product vision, users, goals, constraints, guidelines (optional section) - Read
draft/tech-stack.md— languages, frameworks, patterns, code style, accepted patterns - Read
draft/.ai-context.md(if exists) — system map, modules, data flows, invariants, security architecture. Falls back todraft/architecture.mdfor legacy projects. - Read
draft/workflow.md— TDD preference, commit conventions, review process - Read
draft/guardrails.md(if exists) — hard guardrails, learned conventions, learned anti-patterns - Read
draft/tracks.md— existing tracks to check for overlap or dependencies - Scan recent track impact memory (overlap detection): for each completed track in
draft/tracks/*/metadata.jsonupdated within the last 30 days, read theimpactblock (if present). Build a mapmodule → [recent_track_ids]. After Step 4 (scope distillation), once the candidate modules for the new track are known, intersect them with this map. If overlap exists, surface it in the intake summary:
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.
- 1mo ago First seen · 897 lines · 78 tokens per session scan A b2bfbd9238e5
new-track is a skill published in the GitHub repository drafthq/draft (40 stars, last pushed 12d ago), licensed MIT. It adds 78 tokens to every session and 6,719 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
gentle-ai-collab-perfect
Trigger: contributing to Gentleman-Programming/gentle-ai as an external collaborator. Strict issue-first workflow, honest PR bodies, contributor-vs-maintainer scope, chained-PR strategy, verification protocol, docstring coverage. Load whenever the active repo is Gentleman-Programming/gentle-ai and any part of the…
issue-creation
Trigger: issue creation, bug reports, feature requests, or issue approval. Create and triage GitHub issues from repository evidence.
branch-pr
Create Gentle AI pull requests with issue-first checks. Trigger: creating, opening, or preparing PRs for review.
skill-creator
Trigger: new skills, agent instructions, documenting AI usage patterns. Create LLM-first skills with valid frontmatter.
work-unit-commits
Plan commits as reviewable work units. Trigger: implementation, commit splitting, chained PRs, or keeping tests and docs with code.
cognitive-doc-design
Design docs that reduce cognitive load. Trigger: writing guides, READMEs, RFCs, onboarding, architecture, or review-facing docs.