# Report readiness and open questions

> *AI-generated recommendations for discussion. These are not Nick’s recommendations or approved JLL/PRISM guidance.

Operational status verified September 29, 2026. Competitor research retains its September 13–14 evidence dates. [Current system and limitations](https://crefieldnotes.com/LIVE-MONITORING.md).

## Readiness by use

| Intended use | Assessment | Practical boundary |
|---|---|---|
| Internal strategy discussion | Ready | Use the colleague brief to frame decisions; distinguish evidence from proposed bets. |
| Buyer discovery and demo planning | Ready | Treat reviews and documentation gaps as questions to test in the buyer's configuration. |
| External comparative claims | Conditional | Confirm the offered PRISM configuration and substantiate the particular comparison. |
| Roadmap investment or agent implementation | Requires further evidence | Complete capability validation, customer ground truth and integration scoping before dependent builds. |
| Local agent research | Usable | The package supplies canonical sources, selected claims, linked records, JSONL and a local text-search database. |
| Hosted research and AI retrieval | Live | Public Cloudflare site, 15-tool read-only MCP, daily review checks and weekly broader collection. Current coverage and failures remain visible. |
| Private enterprise integrations | Not connected | No company SSO, CRM, Teams, private PRISM data, automated Confluence sync or vector service is configured. |

## Material research gaps

The gaps below identify evidence required for stronger decisions. They are not findings that either competitor lacks a feature. Proposed owner roles are suggestions for the internal team; no person has been assigned work.

| Gap | Why it matters | Proposed owner | Evidence that closes the gap |
|---|---|---|---|
| Current PRISM capability and entitlement baseline | Public documentation cannot establish the exact sellable configuration or private roadmap. | Product and solution engineering | Dated demonstrations mapped to release, module, license, user role and native/configured/manual/prototype status. |
| Customer value of the proposed financial wedge | A plausible workflow is not demonstrated demand or willingness to pay. | Product, customer success and a controller | Representative labeled records, current reconciliation effort, legitimate omissions, costs and incremental benefit versus the incumbent. |
| Matched competitor execution | Screenshots cannot establish task speed, AI correctness, reliable recovery or successful final accounting state. | Solution engineering and UX research | The same buyer-relevant tasks performed in current authorized environments, with outcomes and human effort recorded. |
| Comparable pricing and total cost | No complete three-vendor quote comparison exists. | Sales operations and finance | Matched modules, AI entitlements, implementation, migration, integrations, support and ongoing operating effort. |
| Deal and buyer evidence | Public positioning does not show which objection or channel actually wins deals. | Sales operations and customer research | Recent comparable win/loss cases, buyer interviews, procurement criteria and reasons for selection; preserve segment and incumbent context. |
| Migration and integration contracts | Integration logos do not prove export completeness, write support or downstream readback. | Integration engineering | Field mappings, permissions, history and attachment coverage, identifiers, retries, limits, reconciliation and accountable support ownership. |
| Procurement requirements | Public trust claims are insufficient for a specific buyer's review. | The relevant security, IT and procurement owners | Current product-specific assurance documents, contractual data handling, access controls and deployment requirements. No compliance conclusion is established here. |
| Addressable demand and competing alternatives | Two rivals do not define the whole market or the size of a profitable segment. | Strategy and finance | Target-account counts, service volumes, budgets and comparison with the incumbent process and relevant adjacent vendors. |

## Presentation and consistency

The evidence library is an appendix for focused investigation. A colleague should begin with the concise [strategy brief](https://crefieldnotes.com/colleague-brief.html), then choose a role-specific route. Opening with 1,334 images or several similarly weighted reports would obscure the decision.

The broad market report and PRISM playbook use different frames. Operational readiness is the general entrant opportunity; financial follow-through is the proposed PRISM experiment. Neither constitutes a demonstrated exclusive market position. The brief reconciles the priority order without inventing a quantitative opportunity score.

Current billing, routing and tenant-app documentation is incorporated into the primary capability matrix. Older reviews remain dated signals with counterevidence. The original 25-page PDF is a historical package snapshot and does not contain the complete expanded library or this readiness layer.

## Knowledge-base limits

The [knowledge-base guide](https://crefieldnotes.com/knowledge-base-guide.html) describes the schema, search, citation rules and update process. The curated claim register covers decision-critical propositions; it is not an exhaustive sentence-level annotation of every report. Longer analysis sections can contain both facts and recommendations, and are labeled accordingly.

Image records are searchable by titles, source captions, roles and linked workflows. Most newly collected figures have generic captions. Their internal text, individual controls, bounding boxes and error states have not all been extracted or semantically annotated. An agent must inspect the image before making a new visual claim. Historical and current illustrations coexist.

The local database uses lexical full-text search. It provides deterministic lookup and provenance; semantic recall, image understanding and downstream model answer quality have not been benchmarked. Retrieval checks validate a bounded set of known questions and reference joins. They do not validate an arbitrary agent or a production RAG service.

## Before external reuse

Select the actual claim and buyer context, verify its supporting evidence and date, and confirm the PRISM capability being offered. Preserve the difference between a vendor assertion, a customer account, documented functionality and a tested result. Recheck mutable pricing, releases and comparative claims before reusing them after this September 13, 2026 snapshot.

The existing agent work pack starts with these dependencies. The knowledge base supports research and scoping; it does not execute those work packages or grant authority to act in product or customer systems. [Agent work pack](https://crefieldnotes.com/prism-agent-work-pack.json)