Skip to content

BdM-15/capture-insights / skills/rfp-reverse-engineer

rfp-reverse-engineer

Verified

Reverse-engineers a federal RFP we received — given the SOW/PWS and evaluation criteria already in the Theseus KG, reconstructs the CO's hidden decision tree (upstream `sow-pws-builder` 6 scope blocks + 3 intake answers), surfaces hot buttons, ghost language, discriminator hooks, missing-section signals, and CPFF-form / Section-5 / QASP / Key-Personnel traps. USE WHEN the user asks "what scope decisions did the CO already make?", "reverse engineer this RFP", "what hot buttons are hiding in this PWS?", "where are the discriminator hooks?", "did they pick CPFF completion or term form?", "anything suspiciously missing?", or any variant of decoding CO intent. Pulls `requirement`, `deliverable`, `proposal_instruction`, `evaluation_factor`, `clause`, `performance_standard` from the active workspace KG and emits a JSON envelope feeding `proposal-generator`. DO NOT USE FOR proposal prose (`proposal-generator`), pricing (`price-to-win`), clause audit (`compliance-auditor`), or sub SOW (`subcontractor-sow-builder`).

references/reverse_engineering_catalog.md

View on GitHub
# Reverse Engineering Catalog — Reading the CO's Hidden Decisions

> **Stance inversion** — Upstream `sow-pws-builder` walks a CO FORWARD through 3 intake answers + 6 scope blocks. This skill walks a bidder BACKWARDS through the resulting document to reconstruct what the CO chose. Same decision tree; opposite direction.

## Section A — Backwards Inference of the 3 Intake Answers

### A.1 SOW vs PWS

| Document Signal                                                                                                     | Inference                                                     |
| ------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------- |
| Section 3 organized by **task areas** with prescriptive verbs ("the contractor shall perform / deliver / maintain") | **SOW**                                                       |
| Section 3 organized by **performance objectives** with outcome verbs + measurable thresholds + AQLs                 | **PWS**                                                       |
| Section 12 titled "Inspection and Acceptance" citing FAR 52.246-series                                              | **SOW** confirmed                                             |
| Section 12 titled "QASP" citing FAR 37.602 / 46.401                                                                 | **PWS** confirmed                                             |
| Mixed signals (task areas + AQLs)                                                                                   | **Hybrid** — emit `status=ambiguous` and ask in clarification |

### A.2 Contract Type

| Clause / Section Signal                                                   | Inference                                                          |
| ------------------------------------------------------------------------- | ------------------------------------------------------------------ |
| Section L pricing schedule asks for **fixed monthly price per CLIN**      | FFP                                                                |
| Section 5 "Labor Category Ceiling Hours" present + FAR 16.601(c)(2) cited | T&M or LH                                                          |
| FAR 52.232-7 cited (Payments under T&M and LH contracts)                  | T&M / LH confirmed                                                 |
| FAR 16.306(d)(1) cited or "completion form" language in Section 1.1       | CPFF completion                                                    |
| FAR 16.306(d)(2) cited or "term form" / "level of effort" language        | CPFF term                                                          |
| Bare `FAR 16.306` cited without (d)(1)/(d)(2)                             | **AMBIGUOUS** — emit `cpff_form_signal.status=ambiguous`, file Q&A |
| FAR 16.305 cited                                                          | CPAF                                                               |
| FAR 16.404 cited                                                          | CPIF                                                               |
| FAR 16.301-3 cited (approved accounting system prereq)                    | Some flavor of CR — confirm via other cites                        |
| Multiple contract types signaled across CLINs                             | Hybrid — list per-CLIN                                             |

### A.3 Commercial vs Non-Commercial

| Signal                                                       | Inference                 |
| ------------------------------------------------------------ | ------------------------- |
| FAR Part 12 cited; FAR 52.212-4 in Section I                 | Commercial                |
| FAR 12.207 cited                                             | Commercial FFP            |
| Full Section I/L clause buildout with FAR Part 15 procedures | Non-commercial            |
| Classified work, weapon systems, government-unique processes | Non-commercial            |
| Both commercial and non-commercial signals                   | Hybrid acquisition — flag |

## Section B — Reverse-Walk of the 6 Scope Blocks

For EACH block, emit a JSON object with `co_choice` (the value visible in the document or inferred), `derivation_signal` (the chunk text or KG entity that fired), `source_chunk_ids` (array), and `status` (`locked` | `implied` | `open`):

- **locked** — visible in the document (verbatim or near-verbatim quote available).
- **implied** — derived via an inference rule firing on visible language.
- **open** — document is silent; emit a `clarification_questions[]` entry.

### Block 1: Mission and Service Model

Walk: Section 1.2 Background, Section 3 task/objective titles, Section 7 Place of Performance.

- `core_service`: extract from background paragraph + recurring noun phrases in Section 3 titles.
- `service_delivery_model`: prime FTE | fully sub-delivered | hybrid — inferred from Key Personnel roles + transition-in language.
- `coverage_model`: business hours | extended | 24/7 | seasonal — inferred from Section 6/7 phrases ("during business hours", "24x7", "follow-the-sun").
- `geographic_scope`: count Section 7 sites; check for "remote" / "virtual" / "telework authorized".

### Block 2: Technical Scope

Walk: Section 3 requirements, Section 8 GFP/GFI, Section 14 constraints.

- `build_vs_buy`: COTS keywords ("commercial off-the-shelf", "SaaS", "configure") vs custom dev keywords ("develop", "design and build").
- `systems_in_scope`: extract every named system; mark new build (Section 3 task) vs existing (Section 8 GFI).
- `integration_complexity`: count distinct system interfaces named in Section 3 + Appendix C.
- `data_migration`: keywords "migrate", "convert", "data transfer from".
- `ai_automation`: keywords "AI", "ML", "NLP", "chatbot", "predictive", "rules-based".

### Block 3: Scale and Volume

Walk: Appendix B (Volume Data), Section 3 throughput requirements.

- `transaction_volume`: extract numbers from Appendix B; if Appendix B absent, mark `open` with high-priority Q&A.
- `user_population`: internal vs external from Section 3 audience phrases.
- `concurrent_users`: extract from performance requirements.
- `growth_projection`: explicit statement vs absence (absence → `open`).

### Block 4: Organizational Scope

Walk: Section 1.2 Background, Section 7 Place of Performance.

- `organizational_units`: count distinct offices/centers/divisions named.
- `phasing`: keywords "pilot", "phased rollout", "all sites at award".
- `stakeholder_complexity`: single program | multi-office | cross-agency.

### Block 5: Contract Structure

Walk: Section 6 Period of Performance, Section 13 Transition.

- `period_of_performance`: extract base + option years.
- `base_year_scope`: full performance | ramp-up | design only — inferred from Section 13.1 transition-in length and Section 6 milestones.
- `clin_structure`: by period | by function | by deliverable — inferred from any pricing schedule visible.
- `transition`: in/out/both — Section 13 subsection presence.

### Block 6: Quality and Oversight

Walk: Section 12 QASP, Section 11 Reporting, Section 10 Key Personnel.

- `aql_thresholds`: extract every AQL stated in Section 12.
- `reporting_cadence`: extract from Section 11.
- `key_personnel`: list every Key Personnel role in Section 10; check substitution clause for the FAR 52.237-2 trap (see `far_citations.md`).

## Section C — Hot Button Decoder

A "hot button" is something the CO emphasized that maps to a discriminator opportunity for our proposal. Detection patterns:

| Document Pattern                                                         | Hot Button Signal                                                   |
| ------------------------------------------------------------------------ | ------------------------------------------------------------------- |
| Threshold elevated above industry baseline (e.g., 99.99% uptime)         | CO has lived through an outage — emphasize reliability proof points |
| Section 13 Transition has unusually detailed risk language               | CO had a bad transition — emphasize transition methodology          |
| Section 12 has multiple "negative incentive" / payment deduction clauses | CO has had quality issues — emphasize QA discipline                 |
| Key Personnel list is unusually long or specific                         | CO values continuity — emphasize team retention strategy            |
| Section 9 Security has agency-specific tier above baseline               | CO has compliance scrutiny — emphasize security posture             |
| Recurring phrase or theme across 3+ sections                             | CO is signaling a priority — emphasize proportionally               |
| Evaluation criteria weight one factor heavily (>40%)                     | That factor is THE buying decision — make it your win theme spine   |

For each hot button, emit `discriminator_opportunity` field linking to the proposal-side response strategy.

## Section C′ — When the catalog is silent

If a document signal does not match any catalog row, do NOT invent a new inference rule on the fly. Emit `status=open` with a `clarification_questions[]` entry describing the signal and asking the user / capture team to confirm. The catalog is the source of truth; updates to it are a separate skill-creator workflow.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall