Skip to content

malekokour/clinpharm-pmx-skills / skills/review-adc-analyte-strategy

review-adc-analyte-strategy

Verified

Reviews the analyte strategy for an antibody-drug conjugate — which analytes are measured, how each is defined, whether the set supports the exposure questions the programme asks, and whether reported parameters are internally consistent across analytes. Use this skill when someone asks to review, QC, or check the PK characterisation plan or results for an ADC or other multi-analyte modality — for example "does this analyte set support a DAR-shift question" or "check that total antibody and conjugated payload are consistently defined across these documents". Do not use to select an analyte strategy, to decide whether an analyte is scientifically necessary, to review bioanalytical method validation, or to review a single-analyte small-molecule programme.

FileKindSize
PASTE.mdmarkdown12.4 KB
README.mdmarkdown2.8 KB
SKILL.mdskill md12.7 KB
references/evidence-hierarchy.mdmarkdown2.7 KB
references/human-review.mdmarkdown2.4 KB
references/output-states.mdmarkdown2.5 KB
references/source-preflight.mdmarkdown2.8 KB
scripts/analyte_consistency.pyexecutablescript7.5 KB

references/output-states.md

View on GitHub
---
contract: output-states
version: "1.0"
owner: Malek Okour
reviewed: "2026-08-04"
consumers: [review-csr-pk-consistency]
---

# Output states

Three markers exist so that missing evidence is never filled by invention. Use
the exact tokens; they are greppable and they appear in evaluations.

| Marker | Means | Use when |
|---|---|---|
| `NEEDS_INPUT` | The check is possible, but a required input is absent | The user can supply the missing artefact and the check will then run |
| `UNKNOWN` | The evidence is present but does not determine an answer | The documents genuinely do not say, and no amount of re-reading will settle it |
| `CANNOT_ASSESS` | The check cannot be performed at all in this configuration | Extraction failed, the format is unsupported, or the check is out of scope for the selected mode |

## Rules

1. **A marker is a result, not a failure.** Emit it plainly, in the same output
   structure as any other finding, with the same location fields. Do not bury it
   in prose and do not apologise for it.

2. **Every marker names what would resolve it.** `NEEDS_INPUT: Table 14.2.4 not
   supplied — dose-proportionality claim cannot be reconciled` is useful.
   `NEEDS_INPUT` alone is not.

3. **Never substitute a plausible value.** If a number is required and absent,
   the answer is a marker, not an estimate, not a typical value, and not a value
   carried over from a similar document.

4. **Never convert a marker into a conclusion.** "No discrepancy found" and
   "could not check" are different results. Reporting the second as the first is
   the most consequential error this contract prevents.

5. **Coverage is stated numerically.** Any output that reconciles values reports
   how many were checked and how many could not be, so a reader can tell the
   difference between a clean document and an unread one. A completeness claim
   without a denominator is unfalsifiable.

## Mechanical finding versus scientific interpretation

A deterministic check produces a **mechanical finding**: two strings differ, a
unit is inconsistent with the stated convention, a ratio does not recompute.

A mechanical finding is never reported as a scientific conclusion. The output
distinguishes them explicitly:

- *mechanical* — "Synopsis states 412 ng·h/mL; Table 14.2.1 states 481 ng·h/mL.
  Values differ."
- *not permitted* — "The synopsis value is wrong." / "This is a minor issue."
  / "The correct value is 481."

Which value is correct, and whether the difference matters, are scientific
judgements reserved for a qualified human reviewer.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 323kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 323kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 323kUpdated Claude CodeCodexVerifiedInstall