AIPOM Context Readiness Advisor
What Is It
Diagnose whether a specific AI-assisted decision or workflow has the trusted, bounded, accessible, and maintainable context it needs. Recommend the smallest intervention that addresses the consequential gap.
Why Use It
Output failures often originate in source ownership, staleness, conflict, access, or context overload rather than model capability. This advisor prevents “add more documents” from becoming the default remedy.
When to Use It
Use before building a context package, connecting retrieval, expanding an agent, or after groundedness and consistency failures.
What It Produces
- Readiness diagnosis by context failure mode
- Evidence, assumptions, and affected decisions
- Priority intervention and owner
- Test and improvement evidence
Who Should Participate
Include the workflow and source owners, intended users, technical implementer, and data, privacy, security, legal, or records partners as needed.
Evidence to Bring
Bring real sources, owners, permissions, refresh history, retrieval traces, failure cases, conflicting decisions, exclusions, and user evidence.
How to Do It
- Define the decision, users, scope, and consequence of bad context.
- Extract sources, owners, access, freshness, conflicts, and known failures.
- Diagnose missing, stale, conflicting, inaccessible, excessive, sensitive, or untrusted context.
- Distinguish document/context problems from underlying data fitness problems.
- Present intervention choices and recommend the smallest useful motion.
- Name owner, test cases, evidence expected, and review trigger.
Facilitation Protocol
Support guided, context-dump, and best-guess modes. Inspect supplied artifacts before questioning. In best-guess mode label authority and freshness assumptions and do not claim organizational readiness from one clean example.
Decision Logic
- Use an authoritative source map when ownership, trust, permissions, or conflict span many sources.
- Build a context package when trusted sources exist but need purpose-bounded assembly.
- Run a data readiness audit when provenance, quality, consent, or representativeness is the root issue.
- Design lifecycle controls when refresh, persistence, expiry, or versioning is failing.
- Narrow or stop the workflow when sensitive or untrusted context cannot be governed.
Completion Criteria
Finish with the failure mode, supporting evidence, affected decision, recommended intervention, owner, representative test, assumptions, and unresolved constraints.
Key Concepts
- Context readiness is purpose-specific.
- Authority, accessibility, and freshness are different properties.
- More context can reduce signal and increase exposure.
- Missing evidence of maintenance caps maturity claims.
Organizational Applications
Use for assistants, retrieval, agents, product discovery, portfolio reviews, support, evaluation, and governance evidence.
Common Pitfalls
- Blaming the model before inspecting sources
- Treating a shared drive as an authoritative system
- Ignoring conflicts and exclusions
- Solving a data problem with prompt text
- Measuring retrieval while ignoring decision quality
Combine With
Use aipom-context-package-builder, authoritative-source-map, or aipom-data-readiness-audit according to the diagnosed root condition.
Assets and Templates
- Readiness template
- Synthetic worked example
- Weak example
Sources
This advisor is an original AIPOM synthesis of context-engineering, knowledge, and data-readiness practices.
1---2name: aipom-context-readiness-advisor3description: Diagnose missing, stale, conflicting, inaccessible, excessive, sensitive, or untrusted context and recommend the next source, package, data, or lifecycle intervention.4---56# AIPOM Context Readiness Advisor78## What Is It910Diagnose whether a specific AI-assisted decision or workflow has the trusted, bounded, accessible, and maintainable context it needs. Recommend the smallest intervention that addresses the consequential gap.1112## Why Use It1314Output failures often originate in source ownership, staleness, conflict, access, or context overload rather than model capability. This advisor prevents “add more documents” from becoming the default remedy.1516## When to Use It1718Use before building a context package, connecting retrieval, expanding an agent, or after groundedness and consistency failures.1920## What It Produces2122- Readiness diagnosis by context failure mode23- Evidence, assumptions, and affected decisions24- Priority intervention and owner25- Test and improvement evidence2627## Who Should Participate2829Include the workflow and source owners, intended users, technical implementer, and data, privacy, security, legal, or records partners as needed.3031## Evidence to Bring3233Bring real sources, owners, permissions, refresh history, retrieval traces, failure cases, conflicting decisions, exclusions, and user evidence.3435## How to Do It36371. Define the decision, users, scope, and consequence of bad context.382. Extract sources, owners, access, freshness, conflicts, and known failures.393. Diagnose missing, stale, conflicting, inaccessible, excessive, sensitive, or untrusted context.404. Distinguish document/context problems from underlying data fitness problems.415. Present intervention choices and recommend the smallest useful motion.426. Name owner, test cases, evidence expected, and review trigger.4344## Facilitation Protocol4546Support guided, context-dump, and best-guess modes. Inspect supplied artifacts before questioning. In best-guess mode label authority and freshness assumptions and do not claim organizational readiness from one clean example.4748## Decision Logic4950- Use an authoritative source map when ownership, trust, permissions, or conflict span many sources.51- Build a context package when trusted sources exist but need purpose-bounded assembly.52- Run a data readiness audit when provenance, quality, consent, or representativeness is the root issue.53- Design lifecycle controls when refresh, persistence, expiry, or versioning is failing.54- Narrow or stop the workflow when sensitive or untrusted context cannot be governed.5556## Completion Criteria5758Finish with the failure mode, supporting evidence, affected decision, recommended intervention, owner, representative test, assumptions, and unresolved constraints.5960## Key Concepts6162- Context readiness is purpose-specific.63- Authority, accessibility, and freshness are different properties.64- More context can reduce signal and increase exposure.65- Missing evidence of maintenance caps maturity claims.6667## Organizational Applications6869Use for assistants, retrieval, agents, product discovery, portfolio reviews, support, evaluation, and governance evidence.7071## Common Pitfalls7273- Blaming the model before inspecting sources74- Treating a shared drive as an authoritative system75- Ignoring conflicts and exclusions76- Solving a data problem with prompt text77- Measuring retrieval while ignoring decision quality7879## Combine With8081Use `aipom-context-package-builder`, `authoritative-source-map`, or `aipom-data-readiness-audit` according to the diagnosed root condition.8283## Assets and Templates8485- [Readiness template](template.md)86- [Synthetic worked example](examples/worked-example.md)87- [Weak example](examples/weak-example.md)8889## Sources9091This advisor is an original AIPOM synthesis of context-engineering, knowledge, and data-readiness practices.
Dean Peters