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 softspark/ai-toolkit --skill research-masterygit clone --depth 1 https://github.com/softspark/ai-toolkitWrote 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/softspark/ai-toolkit/research-mastery)<a href="https://agentmods.dev/skills/softspark/ai-toolkit/research-mastery"><img src="https://agentmods.dev/badge/skills/softspark/ai-toolkit/research-mastery.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00039 | $0.01683 |
| Opus 5 | $0.00019 | $0.00842 |
| Sonnet 5 | $0.00008 | $0.00337 |
| Haiku 4.5 | $0.00004 | $0.00168 |
Grade A, and why
research-mastery 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 5d 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 — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research Mastery Skill
You are not a guessing machine. You are an information retrieval engine.
🔴 The Hierarchy of Truth (Strict Order)
You MUST search in this order. Do not skip steps.
1. Local Knowledge (RAG-MCP)
Source of Truth: The project's Knowledge Base (kb/).
Tool: smart_query(query) (Standard) OR crag_search(query) (High Precision)
Why: This is YOUR project context. It overrides everything else.
Protocol:
- Try
smart_query("task context"). - CRITIC (Self-Correction):
- "Did the docs answer the specific question?"
- If NO: Use
crag_search(query, relevance_threshold=0.7). - If STILL NO:
- LOG GAP: Append query to
kb/gaps.log - Proceed to Step 2.
- LOG GAP: Append query to
2. Context7 (External MCPs)
Source of Truth: Connected MCP servers (e.g., databases, external APIs).
Tool: use_mcp_tool(...)
Why: Live data from the environment.
3. External Search (Internet)
Source of Truth: The Web.
Tool: search_web(query)
Why: For documentation of public libraries not in KB.
Rule: ONLY if Step 1 & 2 yield nothing.
4. Built-in Knowledge (LLM Training)
Source of Truth: Your training data. Why: Fallback for general programming concepts. Rule: Use only for generic syntax/logic, NEVER for project specifics.
🚦 Retrieve-vs-Answer Gate
Before you reach for any tool, decide whether retrieval is even warranted. Two axes settle it:
- Volatility — how fast does the true answer change?
- Timeless / slow-moving (math, definitions, settled algorithms, language syntax): answer directly from built-in knowledge. A search adds latency and noise.
- Current-state / volatile (latest version, today's price, who holds a role now, "is X still the recommended way"): retrieve. Your training data is a snapshot and will lie about the present.
- Recognition — can you place the entity?
- If answering hinges on knowing what some named thing IS (a library, an internal project, a person, an acronym) and you cannot confidently place it, treat that as a signal to search, not to guess. An unfamiliar name is a retrieval trigger, not a hallucination prompt.
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.
- 5d ago First seen · 119 lines · 39 tokens per session scan A e5b589b852f5
research-mastery is a skill published in the GitHub repository softspark/ai-toolkit (170 stars, last pushed yesterday), licensed Apache-2.0. It adds 39 tokens to every session and 1,683 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
cline-fix-volatile-msg
Ladder-aware Cline Anthropic caching — verify the rolling read/write ladder on the wire, then add the tools breakpoint and tune TTL. Updated for the 2026-08 AI-SDK monorepo.
continue-gemini-explicit
Continue's Gemini provider doesn't use the cachedContents API at all. Add explicit caching for sessions over the minimum token threshold.
aider-1h-ttl
Aider uses 5min TTL by default and works around long pauses with keepalive pings. Wire up the 1h TTL beta instead.
cline-openai-cache-key
Cline OpenAI native provider sends no promptcachekey. Add a stable per-task key so cachedtokens stops being zero.
continue-enable-defaults
Continue's prompt caching is opt-in via config and off by default. Flip the default to systemAndTools.
continue-fix-volatile-msg
Ladder-aware Continue Anthropic caching — verify the rolling ladder on the wire, then enable it by default and add TTL coverage.