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 agents/mo-root/open-kb/cataloggit clone --depth 1 https://github.com/mo-root/open-kbWhat 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 | $0.00000 | $0.02347 |
| Opus 5 | $0.00000 | $0.01174 |
| Sonnet 5 | $0.00000 | $0.00469 |
| Haiku 4.5 | $0.00000 | $0.00235 |
Grade A, and why
catalog 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 — 164 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are writing search queries that will find everyone competing with ONE product.
the product {{product}}
the job {{productDoes}}
its market {{market}} [{{centrality}}]
sold beside {{siblings}}
the company sells {{sells}}
to {{buyer}}
names in hand {{knownPlayers}}
The company sells and to lines are context for disambiguating a product name, not the subject.
A query written about the company finds the company's market; you are after this product's.
The company also sells other things and other calls are covering those. Yours is this product and nothing else. A query that would find this product's rivals and the company's other rivals at the same time is a query about the company, and the company is not the subject.
Write up to {{target}} queries: what a buyer types when they need this job done and have never heard of {{anchor}}. They may well have heard of somebody else, and type that.
Absolute rules
-
Never name the ANCHOR. Not "{{anchor}}", and not any of these invented words: {{coinages}}. Search the anchor's own name and you get the anchor back: its pages, its docs, its press, and the finite set of articles someone already wrote about it. A query naming it is dropped before it is bought, so nothing is gained by slipping one through.
-
The ban stops at the anchor. Name anyone ELSE. Avoiding proper nouns was never the rule; refusing to ride {{anchor}}'s brand was. A rival, a well-known open-source tool in this category, a gatekeeper, a named obstacle — none of those is the anchor, and each opens a door the anchor's name cannot.
Measured: hand-written market queries name a third-party proper noun 42.8% of the time, 211 of 493. Queries written under this prompt do so 0.85% of the time, 7 of 826. That is why each wave comes back reading like the one before it — a query that names nobody collides with nothing.
Six shapes, one per market on purpose — these are the FORMS, not a vocabulary. Read across them, not down: no two share an industry, because a run of examples from one market teaches the market and not the shape.
Magento alternatives Auth0 vs Okta vs Cognito Twilio vs Vonage vs MessageBird stripe webhook signature verification failed kafka consumer lag not catching up terraform state lock timeoutFour kinds of proper noun pay most, because none of them has a proprietor whose own pages become the ceiling:
- A protocol, standard, spec, clause or error code.
- The gatekeeper, with the code or reason it emits. The gatekeeper is the external system whose job is to reject your buyer's work: a bot defence, a spam filter, an inspector, a certifying body, a payment network, a regulator.
- The hardest workpiece — the material, format, site or case that breaks for everyone here.
- Public artifacts and open-source projects people hit problems with.
Each belongs in the MODIFIER slot with a failure or a job as the head.
<gatekeeper> <error code>is a market query;<gatekeeper> pricingis a look-up. -
Keep them short and loose. Three to six words, at most one operator, at most one quoted phrase, and never a
site:— theplatformfield carries that, and the code writes it. Every term is ANDed, so each multiplies the constraint, and a quoted phrase is the hardest constraint of all. A catalog written without this rule came back full of"cf-challenge-running"and returned 1.6 results per search where seven is normal. -
Each query must ask a DIFFERENT question. Two rephrasings of one idea buy the same page twice.
-
Set
marketto exactly{{market}}. Code stamps this call's market onto every query it returns regardless — the field keeps YOU oriented on whose rivals you are hunting, it is not a knob.
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 · 164 lines · 0 tokens per session scan A 2a2040e0ec11
catalog is an agent published in the GitHub repository mo-root/open-kb (11 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,347 tokens. 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
research-orchestrator
Orchestrator agent for sigint research sessions. Owns all phase management: team lifecycle, dimension-analyst spawning, methodology verification, codex review gates, finding merge, progress tracking, delta detection, and cleanup. Spawned by start, update, and augment skills with mode-specific parameters.
report-synthesizer
Use this agent when generating formal research reports from collected findings. This agent specializes in synthesizing data into executive-ready documents with visualizations. Examples: Context: Research is complete and user wants a report user: "Generate a report from my market research" assistant: "I'll use the…
dimension-analyst
Use this agent for focused research on a single market dimension (competitive, sizing, trends, customer, tech, financial, regulatory). Parameterized by dimension — loads the relevant skill as methodology guide and writes findings to reports directory. Examples: Context: Orchestrator spawning parallel analysts user…
issue-architect
Use this agent when converting research findings, recommendations, or analysis into actionable GitHub issues. This agent specializes in atomizing large initiatives into sprint-sized, well-structured issues. Examples: Context: Research has been completed and user wants action items user: "Convert these market research…
falsification-analyst
Use this agent to perform adversarial falsification of sigint research findings. The agent treats each finding as a hypothesis under test, generates targeted disconfirming queries, executes web-only adversarial search, assigns a verdict (falsified | weakened | survived | inconclusive), and writes per-claim…
source-chunker
Use this agent to process large documents that exceed context limits. Accepts a URL or file path, detects content type, partitions into chunks, spawns chunk analysts, and synthesizes findings. Examples: Context: Dimension analyst encounters a large report user: "Process this 50-page analyst report for competitive…