# Illustrative output: results-chain alignment

**Fictional example** based on `input.md`. Not a live invocation or validation result.

## Scope and source register

| Source | Type/version/date | Review status |
|---|---|---|
| `input.md#ToC excerpt` | ToC v1.2, May 2026 | Supplied excerpt; approval not stated |
| `input.md#Logframe excerpt` | Logframe v1.0, February 2026 | Supplied excerpt; approval not stated |
| `input.md#Proposal narrative excerpt` | Proposal excerpt, date/version not shown | Supplied excerpt; approval not stated |

## Extraction summary
- **ToC:** Training and follow-up support are expected to build skills (Immediate Outcome), which may lead to method adoption (Outcome) and contribute to harvest resilience (Goal). Seasonal extension visits are an assumption.
- **Logframe:** Farmers adopting at least two methods by end of 2027; indicator wording concerns trained farmers using at least two methods; target 60%.
- **Narrative:** 500 farmers are promised training across North and East districts, while the district counts given total 550. The narrative expects improved harvest resilience.

## Alignment matrix

| Element | ToC | Logframe | Narrative | Status and evidence |
|---|---|---|---|---|
| Training / reach | Training and follow-up build skills. `input.md#ToC excerpt` | Not in supplied excerpt | 500 total promised; district counts 300 + 250 = 550. `input.md#Proposal narrative excerpt` | **Conflict:** narrative total and district sum differ. Not enough evidence to choose a value. |
| Practice adoption | Outcome: some farmers adopt soil-cover methods. `input.md#ToC excerpt` | At least two methods by end of 2027; indicator says “trained farmers using at least two methods.” `input.md#Logframe excerpt` | No adoption target stated. | **Partial:** shared construct, but populations and thresholds are not fully aligned. Indicator definition/denominator/observation window are missing. |
| Harvest resilience | Goal: improved soil retention contributes to resilient harvests. `input.md#ToC excerpt` | No corresponding result in supplied excerpt | Narrative expects improved harvest resilience. `input.md#Proposal narrative excerpt` | **Partial:** broad direction is shared; no measure, period, or pathway detail in the logframe excerpt. This is not evidence of effect. |

## Scope comparison
- **Population/geography:** Logframe indicator uses trained farmers but does not state districts in the excerpt. Narrative names North and East. ToC population is not specified. Alignment is not assessable beyond these statements.
- **Timing:** Logframe target is end of 2027. Proposal says “by 2027” without a precise date. Confirm the intended endpoint.
- **Reach commitment:** Narrative total (500) conflicts with district subtotal (550). Both values remain visible pending an authority decision.

## Indicator identity and version
The logframe's indicator is the only indicator supplied. Its wording is retained exactly. Definition, denominator, and observation window are not supplied, so construct alignment beyond the phrase “at least two methods” cannot be confirmed. No indicator equivalence is inferred from the ToC or narrative.

## Pathway and evidence distinction
- **Hypothesis:** Training and follow-up support build skills; skills may lead to adoption; adoption may contribute to harvest resilience.
- **Stated assumption:** Extension visits occur during planting season.
- **Rationale or cited evidence:** None supplied in these excerpts.
- **Observed result or causal effect:** None supplied. The documents state expected logic, not demonstrated change or attribution.

## Facts, assumptions, and proposed recommendations
- **Facts:** Source wording, versions/dates, target, and counts are as transcribed above.
- **Assumption:** None added to resolve the count discrepancy.
- **Proposed recommendations:** Confirm the intended reach total and district allocation; provide the indicator definition and governing version; clarify the end date and whether harvest resilience is measured.

## Decisions needed
No source was marked governing or approved. The program team should identify which version and commitments control. No automatic authority ranking or rewrite is made.