Individual skill

pho-transportability-audit

Solve Public Health Observatory (PHO) reproducibility/transportability algorithmic-audit tasks. These tasks fetch surveillance + socioeconomic records from a read-only PHO data portal, resolve publication releases/cohorts under a strict evidence specification, run a set of registered statistical audit modules with exact reproducibility checkpoints (fixed seeds, PRNG streams, replicate counts, ordered arrays), apply controlled decision gates, and return one JSON object conforming to a per-task answer template. Invoke when a task references the PHO portal / a "six-module audit" / "transportability" / "publication cohort" / "robustness gates" / analysis_request.json + answer_template.json, or asks to audit whether an exposure-outcome signal survives controlled statistical gates.

experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_01/SKILL.md

Read pinned source

Source commit 5082a777379156a048881e94940d640702e5e09a

What was checked

Matches the observed repository HEAD: Yes
Scanner rules are current: No
Repository scan complete: No
Skill scan complete: Yes
Observed behavior: Not checked yet
Last checked: Nov 6, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings): 121
Repository last updated: Oct 29, 2026
License: Apache-2.0

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

agent Repository topic
ai Repository topic
benchmark Repository topic
llm Repository topic
self-evolving Repository topic
self-learning Repository topic
skills Repository topic
data Derived category
security Derived category

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files

SKILL.mdcore · 9,778 bytes
references/modules.mddoc · 9,839 bytes
references/output_contract.mddoc · 5,378 bytes
references/portal_and_resolution.mddoc · 7,076 bytes

Read excerpt


# PHO Transportability / Algorithmic Audit

## What these tasks are

A PHO audit task gives you three files in `input/`:
- `prompt.txt` — the narrative framing.
- `payloads/analysis_request.json` — **the binding specification**: the business question, geography scope, years, evidence/cohort rules, and a set of named **audit modules** each with an exact `method`, cohort, seeds/streams/replicates, ordered features, grids, and required outputs. It also gives `robustness_gates` / `decision_rule` / `controlled_conclusion`.
- `payloads/answer_template.json` — **the output contract**: required top-level keys, per-section `required_keys`, `array_lengths`, `cardinality_rules`, orderings, enum values, and precision/type rules.

Your job: fetch portal data, resolve it exactly as the spec dictates, implement every module to its registered reproducibility checkpoints, apply the gates, and return **one JSON object** matching the template, with no narrative.

The same five-ingredient structure repeats across the task family (state, county, country scopes; longevity, diabetes, obesity, burden outcomes; FE-OLS, ridge/elastic-net CV, wild cluster bootstrap, split conformal, PCA+k-means, source/year perturbation, GMM mediation, sensitivity surfaces). Master the ingredients, then follow each task's spec literally.

## Entry procedure (do this every time)