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 agentmods add skills/dean0x/devflow/apply-feature-knowledgenpx skills add dean0x/devflow --skill apply-feature-knowledgegit clone --depth 1 https://github.com/dean0x/devflowWhat 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 | $0.00020 | $0.00548 |
| Opus 5 | $0.00010 | $0.00274 |
| Sonnet 5 | $0.00004 | $0.00110 |
| Haiku 4.5 | $0.00002 | $0.00055 |
Grade A, and why
apply-feature-knowledge 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 2d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Apply Feature Knowledge
Iron Law
Pre-computed context, not a cage. Verify against current code — always.
A feature knowledge captures patterns AS THEY WERE when last written. Code evolves. Use the feature knowledge as a starting point, not gospel truth. When something feels off, Read the actual files. Code is authoritative; feature knowledge is supplementary.
3-Step Algorithm
Step 1: Read the Feature Knowledge
When FEATURE_KNOWLEDGE is provided and is not (none):
- Read each feature knowledge section (separated by
--- Feature knowledge: {slug} ---headers) - Absorb: architecture, data flow, key patterns, anti-patterns, gotchas
- Note integration points that relate to your current task
Step 2: Apply to Current Task
- Patterns as defaults: Follow documented patterns unless you have a specific reason not to
- Anti-patterns as warnings: Check your work against documented anti-patterns
- Gotchas as checklists: Verify each gotcha doesn't apply to your changes
- Integration points: Ensure your changes respect documented boundaries
- Key files: Use as starting points for exploration
Step 3: Verify Against Current Code
The feature knowledge may not reflect recent changes:
- If the feature knowledge doesn't address your specific area, explore further
- When an assertion seems outdated: Read the relevant source files to confirm — code wins
- When you find a contradiction between the KB and actual code, trust the code
- Note discrepancies in your output when they matter for the task
Skip Guard
When FEATURE_KNOWLEDGE is (none), empty, or not provided — skip this skill entirely.
Do not mention feature knowledge or its absence in your output.
Freshness Model
Feature knowledge uses write-through + verify-on-read for freshness:
- KBs are written at the point a documented area changes (not on a background schedule)
- Readers verify key assertions against current code rather than relying on staleness markers
- When in doubt, Read the file — that resolves any uncertainty immediately
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.
- 2d ago First seen · 70 lines · 20 tokens per session scan A 584e83392c19
apply-feature-knowledge is a skill published in the GitHub repository dean0x/devflow (19 stars, last pushed 2d ago), licensed MIT. It adds 20 tokens to every session and 548 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
data-engineering
Skill "data-engineering" from fengshao1227/ccg-workflow, covering 数据工程域 · data engineering, 域概览, 数据管道编排, 框架对比 and airflow 核心模式.
verify-change
变更校验关卡。分析代码变更，检测文档同步状态，评估变更影响范围。当用户提到变更检查、文档同步、代码审查、提交前检查、diff分析时使用。在设计级变更、重构完成时自动触发。.
verify-security
安全校验关卡。自动扫描代码安全漏洞，检测危险模式，确保安全决策有文档记录。当用户提到安全扫描、漏洞检测、安全审计、代码安全、OWASP、注入检测、敏感信息泄露时使用。在新建模块、安全相关变更、攻防任务、重构完成时自动触发。.
liquid-glass
Apple Liquid Glass design system. Use when building UI with translucent, depth-aware glass morphism following Apple's design language. Provides CSS tokens, component patterns, dark/light mode, and animation specs.
gen-docs
文档生成器。自动分析模块结构，生成 README.md 和 DESIGN.md 骨架。当用户提到生成文档、创建README、创建DESIGN、文档骨架、文档模板时使用。在新建模块开始时自动触发。.
verify-module
模块完整性校验关卡。扫描目录结构、检测缺失文档、验证代码与文档同步。当用户提到模块校验、文档检查、结构完整性、README检查、DESIGN检查时使用。在新建模块完成时自动触发。.