malekokour/clinpharm-pmx-skills / skills/build-work-context
build-work-context
VerifiedBuilds, updates, refreshes, and exports portable work context for pharmaceutical and biotech professionals. Use this skill whenever someone wants AI to understand their pharma role, department, responsibilities, projects, sources, terminology, constraints, approval rules, or preferred deliverables; asks to be interviewed; mentions a work profile, professional constitution, project instructions, reusable context, Copilot or ChatGPT context, or wants to combine an existing context file with new project documents. Produces a professional context, project capsule, working pack, and change summary without inventing facts. Do not use for patient-specific clinical decisions or as a substitute for qualified scientific, regulatory, or medical judgment.
| File | Kind | Size |
|---|---|---|
| PASTE.md | markdown | 8.2 KB |
| README.md | markdown | 936 B |
| SKILL.md | skill md | 8.4 KB |
| assets/AI-Working-Pack.template.md | markdown | 774 B |
| assets/Context-Change-Summary.template.md | markdown | 494 B |
| assets/My-Pharma-Work-Context.template.md | markdown | 1.7 KB |
| assets/Project-Context.template.md | markdown | 808 B |
| references/context-contract.md | markdown | 3.8 KB |
| references/department-prompts.md | markdown | 3.3 KB |
| references/evidence-hierarchy.md | markdown | 2.7 KB |
| references/human-review.md | markdown | 2.4 KB |
| references/interview-and-modes.md | markdown | 3.8 KB |
| references/output-states.md | markdown | 2.5 KB |
| references/safety-and-evidence.md | markdown | 3.0 KB |
| references/source-preflight.md | markdown | 2.8 KB |
| scripts/context_validation.pyexecutable | script | 6.4 KB |
| scripts/findings.py | script | 4.7 KB |
| scripts/validate_context.pyexecutable | script | 1.7 KB |
references/safety-and-evidence.md
View on GitHub# Safety and evidence contract
Read this reference before processing user-supplied project material.
## Data profiles
### PUBLIC_OR_SYNTHETIC
The content is public, fictional, or explicitly redistributable.
- It may be used in public examples when licensing permits.
- Preserve source links and distinguish synthetic values.
- Do not infer that public availability makes a source authoritative.
### INTERNAL_APPROVED
The user confirms that the current AI environment is approved for the exact
material and task.
- Use only the minimum content needed.
- Do not transfer the material to another service, agent, connector, or tool
unless that destination is also approved.
- Avoid reproducing confidential bodies in logs, filenames, or summaries.
- Keep external communication prepare-only.
### RESTRICTED_DO_NOT_PROCESS
The material includes patient-level data, credentials, sponsor-confidential
content, proprietary submissions, or content not approved for the current AI
environment.
- Stop before reading or transforming it.
- Explain the boundary without asking the user to expose more details.
- Offer a synthetic example, abstract schema, or locally approved route.
### UNKNOWN
The environment or material classification is unclear.
- Ask one short question that will resolve the boundary.
- Do not inspect the source while waiting.
## Authority and evidence
Use this claim discipline:
- `SOURCE FACT`: directly stated in an identified source.
- `USER CONFIRMED`: explicitly provided or confirmed by the user.
- `INFERENCE`: reasoned interpretation from facts.
- `ASSUMPTION`: temporary premise requiring confirmation.
- `UNKNOWN`: missing information.
- `CONFLICT`: incompatible statements not yet resolved.
For each consequential project claim, retain:
- the source filename or user confirmation;
- the relevant date or version;
- whether the claim is direct or inferred;
- contradiction status; and
- the human reviewer or decision owner when known.
Do not use majority vote to resolve incompatible sources. Prefer the user's
declared authority order, then current approved primary sources. Keep a conflict
visible when authority is unclear.
## External actions
Generating a draft does not authorize:
- sending or posting;
- submission or approval;
- account, permission, or connector changes;
- purchase or payment;
- medical, dosing, safety, or regulatory decisions; or
- deletion or irreversible changes.
State the required owner action and leave external execution to the user unless
the current conversation separately authorizes it.
## Human review
The working pack must identify which outputs require review by qualified roles.
Do not invent a named reviewer. Use a role such as `clinical pharmacology lead`
or `regulatory reviewer` until the user confirms a person.
## Sensitive-detail minimization
Prefer:
- role labels over personal names;
- project codes over proprietary asset names;
- aggregate or synthetic examples over raw records;
- source locators over copied document bodies; and
- explicit exclusions over vague claims of safety.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| 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: 472k | Updated | Claude CodeCodex | Verified | Install |
| 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: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| 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 available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |