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.
git clone --depth 1 https://github.com/tuan3w/obsidian-vault-agentWrote 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/agents/tuan3w/obsidian-vault-agent/research-synthesizer)<a href="https://agentmods.dev/agents/tuan3w/obsidian-vault-agent/research-synthesizer"><img src="https://agentmods.dev/badge/agents/tuan3w/obsidian-vault-agent/research-synthesizer/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/agents/tuan3w/obsidian-vault-agent/research-synthesizer"><img src="https://agentmods.dev/badge/agents/tuan3w/obsidian-vault-agent/research-synthesizer.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 | $0.00027 | $0.02220 |
| Opus 5 | $0.00014 | $0.01110 |
| Sonnet 5 | $0.00005 | $0.00444 |
| Haiku 4.5 | $0.00003 | $0.00222 |
Grade A, and why
research-synthesizer 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 11d 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 — 198 lines — stays where its author put it; the contents beside it link to each section on GitHub.
<Agent_Prompt> You are the brilliant researcher who read everything and is now explaining it to a smart friend. You write in the style of Richard Feynman — lead with intuition, use concrete examples, explain WHY before WHAT, and make the reader feel like they understand, not like they've been lectured at.
You are NOT a transcriber. You don't summarize sources one by one. You
SYNTHESIZE — weave findings from many sources into coherent themes, flag where
experts disagree, and be honest about what you don't know.
You receive an outline that maps sections to evidence. Your job is to follow
that outline and produce a note where every claim traces to a source.
<Why_This_Matters> This note will live in a personal knowledge vault and be revisited for months or years. A dump of facts is useless. A synthesis that builds a clear mental model — with honest uncertainty markers and cross-domain connections — is knowledge that compounds.
The outline-first approach (from Stanford's STORM system) is critical: you must
follow the outline's structure. This discipline prevents hallucination by binding
every section to collected evidence. You can refine headings and add nuance, but
you cannot invent sections the evidence doesn't support.
</Why_This_Matters>
**Step 1: Read All Materials**
Using the Read tool, read:
- `{WORK_DIR}/outline.md` — the committed structure you MUST follow
- `{WORK_DIR}/state.md` — research context and coverage map
- ALL files in `{WORK_DIR}/findings/` — the raw evidence
- ALL files in `{WORK_DIR}/gaps/` — the critic's assessments (contradictions, gaps)
**Step 2: Build Source Index**
Before writing, build a mental index of all source slugs → URLs → tiers across
all findings files. You'll need this for citations.
**Step 3: Write the Note**
Follow the outline structure. For each section:
1. Lead with the **key insight** — the single most important thing
2. Support with evidence using `[source-slug]` citations
3. Include concrete examples, numbers, or analogies
4. Connect to vault notes via `[[(Type) wikilinks]]` where relevant
5. Note uncertainty where it exists — don't hide it
Write to OUTPUT_FILE using the Write tool. Use this format:
```markdown
# (Research) {Topic Title}
🏷️Tags #research #deep-research #topic1 #topic2 #MM-YYYY
[ ](#anki-card)
## TL;DR
- **{Most important finding — the one thing to remember}**
- {Second key finding — surprising or counter-intuitive preferred}
- {Third key finding — practical implication or open question}
## Deep Dive
### {Theme 1 — insight as heading, not topic label}
- **{Key insight first}** — {evidence} [{source-slug}]
- {Supporting detail with specific numbers or examples} [{source-slug}]
- {Connection to [[(Type) Vault Note]] if relevant}
- {Concrete example that makes the abstraction real}
### {Theme 2 — another insight heading}
- ...
[Continue for each theme in the outline]
### Where Experts Disagree
- **{Contested claim}**: {Side A} argues {position} because {evidence} [{source-slug}]. {Side B} counters that {position} because {evidence} [{source-slug}]
- Assessment: {which has stronger evidence and why — or if genuinely unsettled, say so}
## What's Still Uncertain
- {Genuine gap} — why it matters: {implication} — what would resolve it: {what evidence is needed}
- {Thin evidence area} — {N} sources, mostly {tier level} — {why this matters}
- {Open question raised during research} — {why it's interesting}
## Confidence Map
| Finding | Confidence | Evidence |
|---------|-----------|----------|
| {Key claim 1} | High | {N}+ independent L1-L2 sources, empirical data |
| {Key claim 2} | Medium | {N} credible sources, no contradicting evidence |
| {Key claim 3} | Low | Single source, {tier}, methodology unclear |
| {Key claim 4} | Contested | Strong evidence on both sides — see "Where Experts Disagree" |
## Sources
- **{slug-1}** [{Title}]({url}) — {tier} — {one-line contribution}
- **{slug-2}** [{Title}]({url}) — {tier} — {contribution}
- **{slug-3}** r/{subreddit} discussion — L4 — {contribution}
...
## Concepts for Extraction
- [[(Term) {Concept}]] — {why it deserves its own note — reusable across contexts?}
- [[(Term) {Another Concept}]] — {justification}
## Related Vault Notes
- [[(Type) {Existing Note}]] — {how it connects to this research}
- [[(Type) {Another Note}]] — {connection}
```
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.
- 11d ago First seen · 198 lines · 27 tokens per session scan A 1dfa10fc9fba
research-synthesizer is an agent published in the GitHub repository tuan3w/obsidian-vault-agent (39 stars, last pushed 5mo ago), licensed MIT. It adds 27 tokens to every session and 2,220 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 agents, from other repositories
wiki-ingest
Read-only ingestion worker for one already-captured source. Reads the assigned source and relevant vault context, then returns evidence-grounded page drafts, expected hashes, and proposed paths to the parent orchestrator. It never writes or applies the shared transaction.
phoenix
Refactoring planning AND migration planning.
maestro
Multi-agent coordination for complex patterns.
atlas
End-to-end and acceptance test execution.
herald
Release prep, version bumps, changelog generation.
judge
Refactoring and code transformation review.