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/qxbyte/pluginhub/continuenpx skills add qxbyte/pluginhub --skill continuegit clone --depth 1 https://github.com/qxbyte/pluginhubWhat 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 | $0.00060 | $0.01448 |
| Opus 5 | $0.00030 | $0.00724 |
| Sonnet 5 | $0.00012 | $0.00290 |
| Haiku 4.5 | $0.00006 | $0.00145 |
Grade A, and why
continue 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 — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/specode:continue — resume an existing spec
Host-tool convention 🔧: tool names in this skill — the
Skilltool (invoke another skill by name),AskUserQuestion(structured multiple-choice question) — are written for Claude-family hosts (Claude Code / CodeBuddy), where naming them directly is the most reliable. On a host that lacks one, use its nearest equivalent (a skill-invocation mechanism / a structured-question tool); with no equivalent, fall back to reading the target skill'sSKILL.mddirectly / plain-text prose. The described behavior is what matters, not the exact tool name.
/specode:continue <slug> loads an existing spec's context and reports where it stands, then stops. It never auto-resumes. This skill owns only the load-and-stop entry + phase inference; the actual pipeline lives in the sibling ../spec/SKILL.md.
Resolver
specode CLIs go through the run.sh wrapper in the plugin's scripts/ directory — relative to this skill that is ../../scripts/; use this skill's base directory to turn the relative path into an absolute one (do not resolve env vars, do not find the cache):
sh ../../scripts/run.sh ../../scripts/resolve_root.py get-root
Flow
- slug is required; if missing → report an error and suggest
/specode:listto find slugs. resolve_root.py get-root(exit 3 not configured → first-time setup; exit 4 configured but unreachable — external drive not mounted / path gone → surface the cause and re-prompt for a path rather than reporting "slug not found"; both per../spec/SKILL.md§specsRoot resolution) → on exit 0, locate<specsRoot>/<slug>/; directory not found → report an error and suggest/specode:list.- Read every fixed doc present, infer the phase per the table below, then report a progress brief and stop — do not auto-resume.
Load-and-stop 🔒: /specode:continue <slug> never auto-resumes. It does exactly three things: (1) locate <specsRoot>/<slug>/ and read every fixed doc present; (2) report a progress brief — slug, inferred phase, per-doc existence, tasks.md checkbox progress (x/N; legacy specs: design.md checkboxes), and what the next action would be; (3) stop and wait for the user's instruction. Only when the user says 继续 (or equivalent) does execution resume from the inferred phase; if the user instead supplies requirement changes, digest them into the affected docs first, then ask whether to resume. The "Resume action" column below describes what happens after the user gives the go-ahead — it is not automatic behavior.
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 · 55 lines · 60 tokens per session scan A f388d85cca06
continue is a skill published in the GitHub repository qxbyte/pluginhub (3 stars, last pushed 27d ago), licensed MIT. It adds 60 tokens to every session and 1,448 once invoked, about $0.0003 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
douyin-favorites-to-knowledge
将用户已授权账号中的抖音视频收藏或用户明确指定的喜欢列表配置并同步到本地 Markdown 或 Obsidian 知识库；默认收藏，只有用户明确说喜欢/点赞才切换来源。首次明确选择推荐的 SiliconFlow转录、本地 Whisper 或不转录。不得绕过登录、访问他人账号或泄露 Cookie 与私密数据。.
knowledge-extractor
Entrevista a un experto de dominio (SME) para extraer su conocimiento técnico tácito y sintetizarlo en una Skill reutilizable. Se activa PROACTIVAMENTE ante 3 situaciones: (1) Frustración — el usuario se queja de código, arquitectura o falta de estándares; (2) Ambigüedad sin Reglas — se pide un refactor profundo pero…
MCP Developer
Skill para desarrollar, mantener y extender el servidor MCP de Obsidian. Incluye patrones de código, arquitectura, testing y gestión de paquetes.
Python Patterns
Buenas prácticas y patrones de desarrollo Python para el proyecto MCP. Incluye estándares de código, patrones arquitectónicos, y convenciones.
Refactoring
Guía para refactorizar código Python de forma segura y efectiva. Incluye técnicas de refactoring, detección de code smells, y mejoras.
Test Runner
Skill para ejecutar y gestionar tests. Incluye patrones de testing, fixtures comunes, y estrategias de debugging de tests fallidos.