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 etoyama/insight-blueprint --skill analysis-journalgit clone --depth 1 https://github.com/etoyama/insight-blueprintWrote 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/etoyama/insight-blueprint/analysis-journal)<a href="https://agentmods.dev/skills/etoyama/insight-blueprint/analysis-journal"><img src="https://agentmods.dev/badge/skills/etoyama/insight-blueprint/analysis-journal/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/etoyama/insight-blueprint/analysis-journal"><img src="https://agentmods.dev/badge/skills/etoyama/insight-blueprint/analysis-journal.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 | $0.00094 | $0.02460 |
| Opus 5 | $0.00047 | $0.01230 |
| Sonnet 5 | $0.00019 | $0.00492 |
| Haiku 4.5 | $0.00009 | $0.00246 |
Grade A, and why
analysis-journal 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 9d 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 — 236 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/analysis-journal — Insight Journal Manager
Records the reasoning process behind an analysis design as a structured event journal. Each event captures a discrete reasoning step — observation, hypothesis, evidence, question, decision, reflection, conclusion, or branch.
Based on Narrative Scaffolding (Huang+ IUI 2026) and Sensemaking Loop (Pirolli & Card) frameworks.
When to Use
- During active analysis: recording what you observed, decided, or discovered
- When choosing an analysis method or package (e.g., "CausalImpact を使う")
- When a new question arises during investigation
- When branching to an alternative hypothesis
When NOT to Use
- Creating a new design from scratch (→ /analysis-design)
- Structured reflection and conclusion (→ /analysis-reflection)
- Registering data sources or knowledge (→ /catalog-register)
Workflow
Step 1: Identify Target Design
If $ARGUMENTS contains a design ID (e.g., "FP-H01"), use it directly.
Otherwise, call list_analysis_designs(status="analyzing") and ask user to select.
Validate with get_analysis_design(design_id).
Step 2: Load or Initialize Journal
Read .insight/designs/{design_id}_journal.yaml using the Read tool.
- If exists: load and show summary (event count, last 3 events, open questions)
- If not exists: create with initial structure using the Write tool:
metadata:
design_id: "{design_id}"
created_at: "{now in ISO 8601 JST}"
updated_at: "{now in ISO 8601 JST}"
events: []
Step 3: Record Events (Interactive Loop)
Listen to user input and classify into InsightType:
| User says... | InsightType | Example content |
|---|---|---|
| データを見たら〜 / 〜が見えた / 〜に気づいた | observe |
"2026年1月以降、売上が前年比15%低下" |
| 〜だと思う / 仮説: / 〜が原因では | hypothesize |
"価格改定(+8%)が売上低下の主因" |
| 〜で確認した / データが示す / 〜のエビデンス | evidence |
"回帰分析でp<0.01の有意な負の相関" |
| 〜が気になる / 〜を調べたい / なぜ〜 | question |
"季節要因を除外できているか?" |
| 〜を使う / 〜で分析する / 手法: | decide |
"CausalImpact で因果効果を推定する" |
| 振り返ると / 見直すと / 反省点 | reflect |
"初期の相関分析では交絡が未考慮だった" |
| 結論: / まとめると / 〜と言える | conclude |
"価格改定は売上低下の主因と確認" |
| 別の仮説 / 分岐したい / fork | branch |
→ Step 4 (Branch Workflow) |
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.
- 9d ago First seen · 236 lines · 94 tokens per session scan A f4db94384dd6
analysis-journal is a skill published in the GitHub repository etoyama/insight-blueprint (0 stars, last pushed 17d ago), licensed MIT. It adds 94 tokens to every session and 2,460 once invoked, about $0.0005 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
rival-search-mcp
Deterministic deep research via RivalSearchMCP. 9 tools: 5-engine web search (DuckDuckGo/Bing/Yahoo/Mojeek/Wikipedia), 9-platform social search (Reddit/HN/StackOverflow/Dev.to/Medium/ProductHunt/Bluesky/Lobste.rs/Lemmy), 5-source news (Google/Bing/Guardian/GDELT/DDG), 5 academic DBs…
safe-email-operations
Use email through the bundled mcp-email-server MCP server, diagnose bounded non-secret state, and hand account or credential setup to a user-operated CLI or authenticated local UI.
naver-search-mcp
Use for Korean web search, Naver News, Blog, Cafe, Image, Knowledge iN, Encyclopedia, Local search, and Naver DataLab search-trend and shopping-insight analysis through the published npm MCP server.
mcp-musescore
Use the ghchen99/mcp-musescore MCP server to inspect, navigate, compose, and edit an open MuseScore score through its QML WebSocket plugin.
ocr-expert
You are connected to OCR-MCP, a FastMCP 3.4+ server with 14 OCR backends. This skill teaches you how to use it effectively.
skills
Generate a complete SVG file from a natural language description using SEP-1577 multi-step sampling.