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/discriminator_hooks.md

View on GitHub
# Discriminator Hooks — Reading the Evaluation Section for Hidden Preferences

> A "discriminator hook" is language in evaluation criteria (UCF Section M or non-UCF equivalent) that signals an unstated CO preference our proposal can exploit. This catalog mirrors Shipley methodology and the Theseus extraction prompt's discriminator definitions.

## Hook Pattern 1: "Most Highly Rated" / "Highest Rated"

When evaluation factors include "most highly rated proposal will demonstrate..." or "highest-rated offerors will...", the CO is **defining the ceiling** for that factor — not just the floor. Plain compliance gets you in the door; matching the "highest rated" language wins the factor.

**Action:** Map every "most highly rated" phrase to a corresponding `discriminator_hooks[]` entry citing the `evaluation_factor` and quoting the ceiling language. Tell the proposal team to write to the ceiling, not the floor.

## Hook Pattern 2: Past-Performance Unlock Phrases

Look for phrases like:

- "relevant and recent past performance"
- "of similar scope, magnitude, and complexity"
- "demonstrating successful performance"
- "with the same or higher security tier"

Each phrase **defines the past-performance qualifier** the CO will apply. If our team's past performance matches, that's a hook. If it doesn't match, that's a competitor advantage we need to neutralize (subcontractor cite, JV partner cite, mitigation strategy).

**Action:** For each unlock phrase, emit `discriminator_hooks[]` entry with `qualifier_phrase` and `our_match_status` (`match` | `partial` | `gap`).

## Hook Pattern 3: Ghost Preferences (Word-for-Word Fingerprints)

The CO often writes evaluation language using exact phrases from a favored vendor's prior proposal or a thought-leadership piece. Detection patterns:

- Unusual jargon or proprietary methodology name appearing in evaluation criteria.
- A phrase that reads like marketing copy rather than government acquisition language.
- A capability described with vendor-specific framing ("engineering excellence framework", "value stream optimization") rather than industry-generic terms.

**Action:** Flag as `ghost_preference` candidates. Cross-reference against known incumbent or favored-vendor materials (out of scope for this skill — handoff to `competitive-intel`).

## Hook Pattern 4: FAB-Chain Triggers

Shipley methodology builds Feature → Advantage → Benefit chains. Trigger phrases in evaluation criteria that justify a FAB chain:

| Phrase                   | Triggers FAB Around                                                 |
| ------------------------ | ------------------------------------------------------------------- |
| "demonstrate ability to" | a capability + proof point + customer-impact statement              |
| "describe approach to"   | a methodology + advantage + measurable benefit                      |
| "explain how"            | a process + advantage + reduced risk / faster outcome               |
| "provide evidence of"    | a past-performance cite + advantage + same-customer-risk-mitigation |

**Action:** For each trigger, emit `discriminator_hooks[]` with `fab_chain_trigger` and recommended Feature/Advantage/Benefit slots for the proposal team to fill.

## Hook Pattern 5: Customer Priority Echoes

When the same priority language appears in **both** background sections (Section 1.2) **and** evaluation criteria, that's a `customer_priority` reinforced as a discriminator. The CO is telling you twice — not by accident.

**Action:** Cross-reference `customer_priority` entities from KG against `evaluation_factor` entities. Emit `discriminator_hooks[]` for each echo with the priority phrase quoted.

## Hook Pattern 6: Pain Point Confessions

Sometimes the CO confesses a `pain_point` directly in background or in a tradeoff narrative ("the prior contract experienced delays in...", "previous transitions were challenging because..."). Each confession is a hook to address head-on in the proposal — usually via a discriminator + proof point pair.

**Action:** Pull `pain_point` entities; for each, emit `discriminator_hooks[]` with `pain_point_text` and `recommended_proposal_response`.

## Hook Pattern 7: Weighted-Factor Spotlight

When evaluation criteria explicitly weight one factor at >40% (e.g., "Technical Approach is significantly more important than Past Performance and Price combined"), that factor IS the decision. Build the win-theme spine around it.

**Action:** Detect explicit weighting language; emit `discriminator_hooks[]` with `dominant_factor` and `recommended_win_theme_spine`.

## What this skill emits, what it does NOT emit

This skill EMITS the hooks (signal detection + capture context). It does NOT write the FAB chains, win themes, executive summaries, or discriminator narratives — that's the `proposal-generator` skill's job. Each `discriminator_hooks[]` entry is a structured handoff into the proposal-generator pipeline.
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