GrantKeeper — Federal Research Grant Compliance
Tracks 14 seeded federal research awards across NIH, NSF, DOE, and DARPA. Per-award budget by Uniform Guidance §200 category, indirect-cost rate math, allowability flags (alcohol / first-class travel / spouse travel), no-cost-extension status, burn-rate vs time-elapsed.
What it is
A grants-management prototype shaped for the Uniform Guidance (2 CFR §200) audit. For every active award: where the budget went by category, what the indirect-cost rate math computes, which lines were spent on unallowable activities, and whether the project is on track or burning too fast / too slow.
What’s in it
- 14 seeded awards across the federal funder mix: NIH R01s, R21s, U54 centers, NSF CAREER, IIS, OAC, CHE, DOE Office of Science, DARPA YFA + BAA. Total committed across the portfolio: ~$25M.
- Real budget categories per Uniform Guidance: personnel (§200.430), fringe (§200.431), supplies (§200.94), travel (§200.474), equipment >$5k (§200.439), contractual subawards (§200.331), other direct (§200.456).
- Indirect-cost rate per award (54-62% reflects realistic federally negotiated F&A rates). Total committed = direct + (direct × IDC).
- Allowability findings — spouse travel on PI conference trip ($8.4k flagged under §200.474), first-class flight exceeding GSA per-diem ($1.28k, §200.474(c)), hosted alcohol at PI symposium ($14k, §200.423). Each flagged with the specific UG citation.
- Burn-rate analysis — compares cumulative spend vs time elapsed. Flags fast burn (>140% of expected), slow burn (<40% past midpoint), and NCE candidates (<90 days left + funds unspent).
- No-cost extension flow — eligible / pending / closed states reflected in the status pill. Final-report-due flag triggers within 120 days of project end.
- Prior-approval triggers (§200.407) — equipment >25% of direct costs surfaces a finding to document sponsor approval.
Why this matters
A1 / A2 / Single Audit findings in the federal research space are mostly the same handful: spent on unallowable categories, missed prior-approval threshold, blew the period of performance without an NCE, IDC math wrong. The shape that catches them is exactly the shape this tool implements — every dollar mapped to a UG category, every category compared to the budget, every flag tied to a regulation.
How it ships
Single HTML file, ~36KB. Zero dependencies. The award catalog, budget math, IDC computation, allowability findings, and status logic are 380 lines of vanilla JavaScript.