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 ddanntheman/GrantForge --skill grant-researchgit clone --depth 1 https://github.com/ddanntheman/GrantForgeWrote 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/ddanntheman/grantforge/grant-research)<a href="https://agentmods.dev/skills/ddanntheman/grantforge/grant-research"><img src="https://agentmods.dev/badge/skills/ddanntheman/grantforge/grant-research.svg" alt="Measured on agentmods" 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 | $0.00073 | $0.00831 |
| Opus 5 | $0.00036 | $0.00415 |
| Sonnet 5 | $0.00015 | $0.00166 |
| Haiku 4.5 | $0.00007 | $0.00083 |
Grade A, and why
grant-research 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 8d 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 — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Grant Research
Build two research files inside the grant workspace before drafting starts: the evidence base that substantiates the narrative, and the funder intelligence that shapes how the narrative is framed. Weak applications assert; winning applications prove.
Requires a grant workspace from grant-intake (or create one if the user is researching ahead of a specific opportunity).
Part 1 — Evidence base (research/evidence.md)
Read the requirements matrix to identify what must be proven: typically the need (problem severity in the service area), the intervention's effectiveness, and the org's capacity to deliver.
Need data. Search for the most local, most recent data available, in this order of preference: the org's own tracked data (from the profile), then county/city-level government data, then state, then national. Standard sources by domain: Census/ACS (demographics, poverty), BLS (employment), CDC WONDER and county health rankings (health), NCES and state report cards (education), HUD and NLIHC (housing), FBI UCR and local dashboards (safety), Feeding America Map the Meal Gap (food security). Record for every statistic: the figure, year, geography, source name, and a working URL. Verify each URL resolves before recording it.
Intervention evidence. Search for peer-reviewed studies, meta-analyses, and clearinghouse ratings (What Works Clearinghouse, Blueprints, evidence-based registries relevant to the domain) supporting the program model. Prefer studies of the same model; otherwise the closest analog, stated honestly as an analog.
Rules. Never fabricate or approximate a statistic. If the data does not exist at the needed geography, say so and recommend the honest fallback (state-level data plus local qualitative evidence). Flag any statistic older than five years. Every entry gets a full citation in the format the funder requires (or a consistent format if unspecified).
Part 2 — Funder intelligence (research/funder.md)
Foundations. Pull the funder's 990-PF/990 (ProPublica Nonprofit Explorer, Candid) for the last 2-3 years: total giving, grant count, typical award size, and — most useful — the actual grantee list. Identify grantees similar to the org (size, geography, program). Read the funder's website, annual report, and any president's letter for stated strategy and vocabulary. Note board members and staff with any connection to the org's network.
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.
- 8d ago First seen · 79 lines · 73 tokens per session scan A 6efdb4a0d412
grant-research is a skill published in the GitHub repository ddanntheman/GrantForge (2 stars, last pushed 2mo ago), licensed MIT. It adds 73 tokens to every session and 831 once invoked, about $0.0004 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
inno-grant-proposal
A writing aid for preparing and improving research grant proposals for US and Chinese funding agencies.
Grant Writer
Craft compelling grant proposals that secure funding for nonprofits and research.
kami-deck
A lab-meeting deck on gut-microbiome links to sleep quality — the design, the results, the caveats, and the next experiment. Built as a decision-grade academic research deck for lab group, PI.
watch
File sentinel that monitors the working directory for changes and marker comments, then auto-triggers appropriate skills. Poll-based via git diff against the last scan commit. Writes intake items for batch processing and routes marker actions through /do. Use for automatic reactions to file changes; do NOT use for…
review
5-pass structured code review — correctness, security, performance, readability, consistency.
live-preview
Mid-build visual verification loop. Takes screenshots of components during construction, not just after. Catches visual regressions and invisible features before they compound. Requires Playwright or similar screenshot tool.