K-Dense-AI/scientific-agent-skills / skills/research-grants
research-grants
VerifiedSupports research proposal preparation and review for NSF, NIH, DOE, DARPA, and Taiwan NSTC, including opportunity-specific requirements, aims, review criteria, budgets, broader impacts, forms, and resubmissions. Use for investigator-authored grant development, compliance matrices, and proposal critiques.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 16.9 KB |
| assets/budget_justification_template.md | markdown | 7.2 KB |
| assets/nih_specific_aims_template.md | markdown | 3.0 KB |
| assets/nsf_project_summary_template.md | markdown | 2.5 KB |
| references/README.md | markdown | 2.4 KB |
| references/broader_impacts.md | markdown | 23.8 KB |
| references/core_components.md | markdown | 16.1 KB |
| references/darpa_guidelines.md | markdown | 7.0 KB |
| references/doe_guidelines.md | markdown | 7.5 KB |
| references/nih_guidelines.md | markdown | 13.3 KB |
| references/nsf_guidelines.md | markdown | 7.8 KB |
| references/nstc_guidelines.md | markdown | 7.5 KB |
| references/proposal_types_and_resubmission.md | markdown | 3.2 KB |
| references/review_criteria.md | markdown | 2.8 KB |
| references/specific_aims_guide.md | markdown | 20.6 KB |
| references/writing_principles.md | markdown | 3.5 KB |
references/doe_guidelines.md
View on GitHub# DOE proposal guidance
Reviewed 2026-10-01. DOE does not have one proposal format, review weighting, cost-share
rate, deadline, or submission portal. Start from the current office's NOFO and all
amendments. Distinguish a funding opportunity from an award announcement, RFI, user
facility allocation, procurement, or internal national-laboratory competition.
## Select the program and route
| Route | What to establish before drafting |
| --- | --- |
| Office of Science (SC) | Scientific topic fit, program contact, pre-application/LOI requirements, full-application route, and review criteria. |
| ARPA-E | Transformative energy objective, concept-paper/full-application stages, measurable targets, cost share and technology-to-market plan. |
| Applied-energy office | Current sponsoring office, topical scope, project stage, deployment/industry requirements and the exact Exchange portal. |
| National lab/user facility | Whether this is a financial-assistance application, lab work authorization, or separate facility-access proposal. |
Use [SC funding listings](https://science.osti.gov/grants/FOAs/Open) and the
[closed-call archive](https://science.osti.gov/grants/FOAs/Closed) to distinguish status.
The FY2026 Early Career Research Program is a dated competition, not a permanently
open $750,000 award. Copy eligibility and budget from the relevant year's notice.
DOE announced [CMEI program realignment in January 2026](https://www.energy.gov/cmei/articles/energy-department-announces-realignment-critical-minerals-and-energy-innovation).
Old EERE/FECM office lists should not determine current application routing. The legacy
`eere-exchange.energy.gov` domain currently identifies **CMEI eXCHANGE**. Use the
sponsoring office and portal in the actual NOFO; historical EERE names may still occur
in authorizing documents and opportunity records.
### Portal contracts
- [Grants.gov](https://www.grants.gov/) hosts federal opportunities and application
packages. Match the exact opportunity and package; check amendments and status.
- [PAMS](https://pamspublic.science.energy.gov/) is SC's management system. Do not send
an ordinary SC full application there merely because pre-applications use PAMS.
- A verified **historical example**, [DE-FOA-0003600 Amendment 000001](https://science.osti.gov/bes/-/media/grants/pdf/foas/2026/DE-FOA-0003600-000001.pdf),
sends pre-applications/LOIs to PAMS and full applications to Grants.gov, not FedConnect.
This illustrates the distinction; it does not establish current availability.
- [ARPA-E eXCHANGE](https://arpa-e-foa.energy.gov/) is the application portal;
**OPEN is a program**, not the name of a submission system.
- [CMEI eXCHANGE](https://eere-exchange.energy.gov/) carries opportunity-specific
application instructions. Do not assume its stage or deadline matches ARPA-E.
These are authenticated user workflows, not documented REST endpoints in this skill.
Public listing checks do not validate account access, institutional roles, uploads,
or successful submission. Establish SAM/UEI, Grants.gov and any agency account/role
requirements early; do not promise immediate registration or transmit credentials.
## Build a compliance matrix
Record each requirement with its NOFO section: eligibility, topic, LOI/concept paper,
encouragement versus invitation, application components, page limits, fonts, file
formats, budget ceiling, period, cost share, review factors, contacts, and deadline
with time zone. Concept-paper feedback is not universally an invitation prerequisite;
the NOFO determines whether an applicant may proceed after discouragement.
For a narrative outline, allocate space only after checking the actual limit:
1. **Problem and objectives:** quantify the scientific or technical barrier and why
the selected program should support the work.
2. **Innovation and evidence:** compare to defensible baselines; separate observed
results from proposed targets and explain key uncertainties.
3. **Technical plan:** map tasks to methods, resources, dependencies, risks,
alternatives, outputs and measurable success criteria.
4. **Impact:** for basic science, state the knowledge gain; for technology programs,
state the performance/cost basis and credible path to users or markets.
5. **Team and execution:** name roles, facilities, access agreements, milestones and
management responsibilities; a national-lab affiliation is not evidence of access.
Use a work breakdown structure when requested. Match task identifiers across the
narrative, schedule and budget. A TRL starting/target level is appropriate when the
NOFO uses TRLs; do not impose fixed TRL ranges on every SC or ARPA-E proposal.
## Personnel forms and sharing
For opportunities published on/after January 1, 2026, covered individuals generate
DOE Common Forms through SciENcv under [DOE's Common Forms guidance](https://www.energy.gov/ia/common-forms-sciencv).
Use the agency's current biosketch and Current and Pending (Other) Support instructions,
certifications and disclosures. An old NSF-style CV is not a safe substitute. Older
opportunities and specific programs can have transitional/additional instructions.
Follow the NOFO's data-management/public-access requirements. Plan for repositories,
formats, metadata, preservation, access restrictions, and costs; do not treat a data
plan as optional merely because an old example omitted it. Check human/animal research,
research security, foreign participation, and other applicable requirements using the
opportunity and [SC policy guidance](https://science.osti.gov/grants/Policy-and-Guidance).
## Budget and cost share
- Justify personnel effort/rates, equipment, supplies, travel, subcontract/subaward
roles, and the applicable indirect-cost rate/base. Match the award instrument.
- Copy the required share and exceptions from the NOFO. There is no general
"20% university/50% industry" ARPA-E rule. Team composition, work allocation,
recipient type and award instrument may change the required share. A dated
[MAGNITO FAQ](https://arpa-e.energy.gov/sites/default/files/2025-11/DE-FOA-0003590%20MAGNITO%20FOA%20FAQ%20-%204f.pdf)
illustrates why reductions must be evaluated for the whole project team.
- If share is `r` of **total project cost**, nonfederal share is
`federal_request * r / (1 - r)`, not `federal_request * r`. Synthetic example:
$800,000 federal at 20% total-project share needs $200,000 nonfederal support.
- Identify the source, valuation, authority, timing and evidence for each permitted
contribution. Do not count the same expense twice or assume another federal award
qualifies. For basic SC research, cost share is often not required; confirm the call.
- Lab participation may use separate work authorizations or other arrangements.
Confirm the allowable funding route with the lab and NOFO rather than assuming a
university-to-lab subcontract. Facility time normally has its own access process.
## Review and submission
SC reviews scientific/technical merit, method, personnel/facilities and budget under
its governing guidance; mission/program relevance and other selection factors also
matter. Copy the actual order or weights from the NOFO. Do not invent percentages
or transplant one ARPA-E program's rubric to another.
Before submission, reconcile all versions and cost-share letters, complete required
institutional review, inspect the generated application, and record accepted receipt
and tracking identifiers. A selection for negotiation is not an issued award. Public
listings, historical success rates, or informal program feedback cannot establish
fundability, authorization to spend, or a guaranteed award date.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |