> ## Documentation Index
> Fetch the complete documentation index at: https://foreverbetter.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Action plans

> Turn an analysis into a prioritized set of lifestyle interventions and an evidence-graded supplement stack, cross-referenced against what the user already takes.

`GET /analyses/{id}/action-plan` builds a customized plan from a stored analysis.
Each out-of-range finding is mapped to lifestyle interventions and supplements,
aggregated across findings, and cross-referenced against the user's logged
supplements and medications.

```bash theme={null}
curl -s "$FB_API/analyses/$AN/action-plan" \
  -H "authorization: Bearer $FB_KEY" | jq .
```

## What you get back

```json theme={null}
{
  "analysis_id": "analysis_...",
  "summary": "4 findings outside target mapped to 6 lifestyle changes and 5 supplements, 2 of them core priorities...",
  "interventions": [
    {
      "id": "reduce_sat_fat",
      "name": "Cut saturated fat, raise unsaturated",
      "category": "nutrition",
      "detail": "Swap butter and fatty processed meat for olive oil, nuts, and fatty fish.",
      "evidence": "A",
      "priority": "core",
      "targets": [{ "marker": "apob", "finding": "ApoB", "direction": "high", "status": "needs_attention" }],
      "rationale": "Targets high ApoB, high Non-HDL-C."
    }
  ],
  "supplements": [
    {
      "id": "omega_3",
      "name": "Omega-3 (EPA/DHA)",
      "dose_guidance": "Exact dose and timing are withheld in the wellness API. Confirm need, product, dose, timing, interactions, and a retest plan with a qualified clinician or pharmacist.",
      "evidence": "A",
      "evidence_note": "Consistently lowers triglycerides; supports inflammation and blood pressure.",
      "priority": "core",
      "targets": [ { "marker": "triglycerides", "finding": "Triglycerides", "direction": "high" } ],
      "rationale": "Targets high Triglycerides, high ApoB.",
      "cautions": ["You logged warfarin: Omega-3 can increase bleeding risk with blood thinners."],
      "already_taking": true
    }
  ],
  "cautions": ["You logged 1 medication with pharmacogenomic relevance (warfarin)..."],
  "evidence_key": { "A": "Strong, consistent human evidence", "B": "Moderate evidence from good trials", "C": "Limited or preliminary evidence", "D": "Mixed or mechanistic evidence" },
  "disclaimer": "Educational wellness guidance, not medical advice..."
}
```

## How it is built

* **Mapped from findings.** Only findings outside their target range drive the
  plan. Each maps by marker and direction (e.g. high ApoB, low vitamin D) to a
  curated set of interventions and supplements.
* **Aggregated.** A supplement or intervention that helps several flagged markers
  is listed once, with every marker it addresses in `targets`, and ranks higher.
* **Prioritized.** `priority` is `core` when it addresses a `needs_attention`
  finding, otherwise `optimize`. Lifestyle interventions lead; supplements follow.
* **Evidence-graded.** Every item carries an A–D `evidence` grade (see
  `evidence_key`) and a short `evidence_note`.
* **Dose-safe by default.** Public wellness responses omit exact supplement dose
  and timing. A deployment can expose them only after enabling a separately
  reviewed clinical policy; consumer UIs must not infer a dose when absent.
* **Personalized against what they take.** Supplements the user already logs are
  marked `already_taking` instead of re-recommended. Known supplement–drug
  interactions surface as `cautions` against the user's medication list, and
  medications with pharmacogenomic (CPIC) relevance surface at the plan level.

## Sources

Each supplement carries a `sources` array, and the plan carries a top-level
`sources` list, so every recommendation can be traced to its evidence:

* **[SUPP.AI](https://supp.ai)** (Allen Institute for AI): supplement–drug
  interaction evidence extracted from the scientific literature. When a user's
  logged medication matches a documented interaction, the caution names the drug,
  the number of supporting literature reports, and links to the evidence.
* **[Pillser](https://pillser.com/health-outcomes)**: per-outcome study counts.
  When a supplement targets a marker with a mapped outcome (e.g. reduced LDL,
  improved insulin sensitivity), the plan cites that outcome page as an evidence
  base.
* The mappings from findings to interventions and supplements, and the A–D
  grades themselves, are the API's own curated knowledge base, informed by the
  sources above.

Interaction cautions are only shown for drugs the user actually logs, so the
underlying interaction dataset never produces noise for medications they don't take.

## Safety

Action plans are educational wellness guidance, not medical advice. They never
diagnose or treat disease and never tell a user to start, stop, or change a
prescribed medication. Confirm nutrient deficiencies with testing, and review
supplements and interactions with a clinician or pharmacist. The `disclaimer`
field carries this language for display.