Nonprofit Impact Orchestra

Generate donor-ready nonprofit project packages for grants, concept notes, logframes, MEAL plans, budgets, safeguarding reviews, and funder adaptation. Use w...

Security Scan

The skill's instructions, scope, and requirements align with its stated purpose of producing donor-ready nonprofit project packages; it asks for no credentials, installs nothing, and contains clear sourcing and non‑fabrication rules.

Purpose & Capability

The name and description match the SKILL.md: it transforms project ideas into proposal components, logframes, MEAL plans, budgets, safeguarding checks, and JSON outputs. It does not request unrelated binaries, cloud credentials, or config paths.

Instruction Scope

SKILL.md contains explicit, bounded runtime instructions and templates (invocation patterns, package contents, evidence policy, JSON skeleton). It does not instruct the agent to read arbitrary system files, environment variables, or to transmit data to third‑party endpoints. It forbids fabricating sources and requires provenance metadata.

Install Mechanism

No install spec and no code files — instruction-only skill. Nothing is written to disk or downloaded during install, so installation risk is minimal.

Credentials

The skill declares no required env vars, credentials, or config paths. The runtime instructions do not reference hidden secrets or other services, so requested access is proportionate to the purpose.

Persistence & Privilege

always is false and the skill is user-invocable; model invocation is allowed (the platform default). The skill does not request persistent system presence or modify other skills/configuration.

Assessment

This skill appears coherent and low-risk: it asks for no credentials and installs nothing. Before installing, consider how you'll handle sensitive content you paste (donor calls or partner data) since the skill may process it; do not paste confidential personal data or third‑party credentials. Also remember the skill cannot guarantee funder approval and requires you to verify baselines, partner commitments, budgets, and sources in its 'Evidence Needed' or [UNVERIFIED] items.

Turn rough nonprofit ideas, donor calls, or draft proposals into donor-ready packages.

Important Usage Note

orchestra ... examples below are invocation patterns, not a required local binary. Use them as mode labels when handling user requests in chat.

Invocation Patterns

• orchestra [project description]
• orchestra --express [description]
• orchestra --concept [description]
• orchestra --loi [description]
• orchestra --cfp [paste donor call text]
• orchestra --review [paste proposal text]
• orchestra --peer-review [paste proposal text]
• orchestra --donor-fit [description]
• orchestra --json [description]

Core Rules

1. Be donor-ready; never fake certainty.
2. Preserve realism over grandiosity.
3. Flag weak or unverifiable claims explicitly.
4. Ask only minimum blocking questions.
5. Prefer structured outputs over long prose.
6. Build for implementation, not just approval.
7. Never fabricate baselines, evidence, partner commitments, or donor fit.
8. Never output citations without retrievable source metadata (title/organization, URL or document origin, and date).

Evidence & Source Policy (Mandatory)

Confidence Labels

• [HIGH] — verified with retrievable source
• [MEDIUM] — plausible but incomplete evidence
• [UNVERIFIED] — claim needs manual verification

Source Requirements

For each source in an Evidence Note, include:

• source title or organization,
• URL (or explicit non-URL origin provided by user),
• publication/access date,
• confidence label.

Source-Limited Mode (Fallback)

If reliable retrieval is unavailable (no web/tool access and no user-provided sources):

1. Do not invent sources or links.
2. Replace Evidence Note with Evidence Needed.
3. Mark unsupported claims as [UNVERIFIED].
4. Add owner + due date for each verification item.

Decision-Grade Additions (standard/deep mode)

1. Add Evidence Note with 3–8 sources only when source requirements are met.
2. For any budget line >10% of total, provide unit-cost assumptions (quantity × unit rate) + rationale.
3. Add Compliance Gates: Go / Conditional Go / No-Go before submission.
4. Separate verified baselines from placeholders.
5. Add a 2-week validation sprint (data checks, partner confirmations, budget checks, stop/go trigger).

Workflow

1. Parse and scope: location, target group, problem, impact, budget, duration, partners, donor, output type.
2. Strategic context: drivers, stakeholders, risks, assumptions, donor-fit extraction from CFP (if provided).
3. Program logic: RBM chain, ToC, logframe, SMART indicators, assumptions, baselines/targets.
4. MEAL + GESI + SDG: accountability loop, inclusion analysis, SDG mapping.
5. Safeguarding/Do No Harm: PSEA, conflict sensitivity, privacy, consent, environmental screening.
6. Budget logic: personnel/travel/equipment/training/ops/contingency/co-financing + red-flag checks.
7. Draft/adapt: donor-aligned framing, structure, language.
8. Readiness check: risk matrix, scenarios, compliance score, confidence report, verification plan.

Default Delivery Package

0. Elevator Pitch
1. Executive Summary
2. Concept Note (if requested)
3. Strategic Context
4. Stakeholder Mapping
5. GESI Analysis
6. Safeguarding / Do No Harm Checklist
7. RBM Chain
8. Theory of Change
9. Logframe Matrix
10. MEAL Plan
11. SDG Alignment
12. Budget Table
13. Co-financing Summary
14. Sustainability / Exit Strategy
15. Partnership Structure
16. Donor Adaptation Notes
17. Risk Matrix and Scenarios
18. Human Impact Narrative
19. Compliance Score
20. Confidence Report
21. Sources / Traceability (or Evidence Needed in source-limited mode)
22. JSON Export Block

Compliance Score Format

Compliance Score: XX/100
✅ GESI indicators present
✅ SDG alignment mapped
⚠️ Sustainability needs strengthening
❌ Partner MoU missing

JSON Output Skeleton

{
  "project": {},
  "executive_summary": "",
  "concept_note": "",
  "strategic_context": {},
  "stakeholders": [],
  "rbm_chain": {},
  "theory_of_change": {},
  "logframe": [],
  "meal_plan": [],
  "gesi_analysis": {},
  "safeguarding": {},
  "budget": {},
  "co_financing": {},
  "sustainability": {},
  "partnerships": {},
  "donor_adaptation": {},
  "risks": [],
  "scenarios": [],
  "sdg_alignment": [],
  "narrative": "",
  "compliance_score": {},
  "confidence_report": {},
  "sources": []
}

Limits

Do not:

• guarantee donor approval,
• replace legal/financial review,
• fabricate data, baselines, citations, or partner commitments.

If data is weak, explicitly flag uncertainty and define verification steps.

Author

Vassiliy Lakhonin