Skip to content

BdM-15/capture-insights / skills/subcontractor-sow-builder

subcontractor-sow-builder

Verified

Drafts a federally-defensible SOW or PWS the prime issues to a subcontractor / teaming partner — same FAR 37.102(d) / 37.602 / 16.601(c)(2) / 16.306(d) discipline a contracting officer applies, opposite seat. USE WHEN the user asks to "write a SOW for our sub", "draft a PWS for [Partner]", "build the teaming-partner statement of work", "convert this SOO into a sub SOW", "we need a SOW the sub will sign", or any variant of authoring a downstream work statement. Walks the upstream 3-phase tree (acquisition intake → 6 scope blocks → 14-section assembly), pulls scope from the active Theseus KG (requirements, deliverables, work_scope_items, performance_standards), enforces FAR 37.102(d) "no FTEs in body", emits a chat-only staffing handoff for the prime's cost build, writes Markdown for `renderers` → .docx. DO NOT USE FOR prime proposal prose (`proposal-generator`), reverse-engineering an RFP (`rfp-reverse-engineer`), pricing the sub (`price-to-win`), or clause audit (`compliance-auditor`).

FileKindSize
SKILL.mdskill md16.3 KB
UPSTREAM.mdmarkdown4.7 KB
evals/evals.jsonconfig5.9 KB
references/decision_tree_blocks.mdmarkdown7.6 KB
references/far_citations.mdmarkdown9.4 KB
references/language_rules.mdmarkdown3.2 KB
references/sow_pws_section_structure.mdmarkdown10.1 KB

evals/evals.json

View on GitHub
{
  "schema_version": "1.0",
  "skill": "subcontractor-sow-builder",
  "evals": [
    {
      "id": "sub-sow-001",
      "name": "Tier-2 cyber engineering PWS for teaming partner",
      "prompt": "We're prime on a DoD task order and need a PWS for our Tier-2 cyber engineering subcontractor. Workspace `afcap6_drfp` has the requirements. Sub will provide threat hunting and detection engineering. Contract type with sub will be CPFF term form (R&D-flavored work, level of effort). Draft the PWS.",
      "expected_signals": [
        "calls `kg_entities` for at least requirement, deliverable, work_scope_item, performance_standard",
        "asks the 3 intake questions (SOW vs PWS, contract type, commercial vs non-commercial) up front in a single batched message",
        "Section 1.1 explicitly cites FAR 16.306(d)(2) (CPFF term form) — never bare `FAR 16.306`",
        "Section 5 is OMITTED (CPFF, not T&M/LH); subsequent sections renumber sequentially",
        "Section 9 references DD Form 254 since classified-adjacent work is implied",
        "Section 10 Key Personnel substitution clause is NOT FAR 52.237-2; agency-appropriate clause or FAR 52.237-3 fallback",
        "Section 12 QASP Payment Consequence column lists invoice deductions / cure notices / re-performance — NEVER CPARS rating language",
        "NO FTE counts, SOC codes, hours-per-year, or staffing tables anywhere in the document body or appendices",
        "After the artifact write, BOTH chat-only handoffs are emitted (Staffing Handoff for `price-to-win`, CLIN Handoff for prime's subagreement Section B)",
        "Markdown artifact saved to `{run_dir}/artifacts/sub_sow_pws.md` via `write_file`"
      ],
      "anti_signals": [
        "Section 5 included for a CPFF contract",
        "Bare `FAR 16.306` citation without (d)(1) or (d)(2)",
        "FTE counts or labor categories appearing in the document body",
        "QASP Payment Consequence column containing 'Satisfactory', 'Exceptional', 'Marginal', 'Unsatisfactory', or any CPARS rating",
        "FAR 52.237-2 cited for Key Personnel substitution",
        "Staffing handoff written to a file or embedded as an appendix"
      ]
    },
    {
      "id": "sub-sow-002",
      "name": "Commercial FFP services SOW for sub under prime BPA",
      "prompt": "We hold a GSA MAS BPA. Need to issue an FFP commercial-services SOW to our help-desk subcontractor for a 1-year base + 4 option years. Workspace `mcpp_drfp` has the prime requirements. Draft the SOW.",
      "expected_signals": [
        "intake answers: SOW (not PWS), FFP, commercial → Section 14 flags FAR Part 12 applicability and FAR 12.207 (FFP commercial)",
        "Section 5 OMITTED (FFP); subsequent sections renumbered",
        "language uses prescriptive SOW verbs ('the contractor shall perform', 'the contractor shall provide') — NOT outcome-oriented PWS verbs",
        "Section 9 cites FAR 52.204-21 (Basic Safeguarding) since CUI is plausibly in scope",
        "CLIN Handoff Table emitted in chat with separate travel and ODC CLINs if travel/ODCs in scope"
      ],
      "anti_signals": [
        "PWS-style outcome verbs in Section 3 of an SOW",
        "Section 5 included for FFP",
        "Travel or ODCs blended into a single performance CLIN"
      ]
    },
    {
      "id": "sub-sow-003",
      "name": "T&M sub SOW with mandatory Section 5 ceiling-hours table",
      "prompt": "We're prime on a follow-on; sub agreement with our incumbent partner will be T&M for surge engineering support. Draft the sub SOW. Performance is at Huntsville, AL. Base + 2 option years.",
      "expected_signals": [
        "intake answers: SOW or PWS user-chosen; T&M; commercial-status user-chosen",
        "Section 5 Labor Category Ceiling Hours table PRESENT, citing FAR 16.601(c)(2)",
        "Section 5 columns: `Labor Category | Base Year Ceiling Hours | OY1 | OY2 | Total Ceiling` — NO SOC codes, NO FTE counts, NO derivation basis",
        "skill recognizes Section 5 is a pricing risk-sharing mechanism, NOT a staffing prescription",
        "Staffing Handoff Table (chat-only) is still emitted separately — never confused with the Section 5 table"
      ],
      "anti_signals": [
        "Section 5 table containing SOC codes or FTE counts",
        "Section 5 omitted from a T&M contract",
        "Bare `FAR 16.601` citation without `(c)(2)`",
        "Section 5 table merged with the chat-only Staffing Handoff"
      ]
    },
    {
      "id": "sub-sow-004",
      "name": "Grounding-audit floor (cover-note skill)",
      "prompt": "Same as sub-sow-001 — issue the standard chat narrative summary that walks the 6 scope blocks and the 14-section assembly, then disclose the chat-only staffing handoff for price-to-win.",
      "expected_signals": [
        "The skill MUST call write_file with target path artifacts/sub_pws_*.md before drafting the chat narrative summary.",
        "tools/audit_skill_grounding.py --skill subcontractor-sow-builder reports grounding_ratio >= 0.85 (insight-heavy 6-scope-block + 14-section reasoning skill; the on-disk PWS Markdown is the citation-of-record, the chat narrative is the human-readable digest; 147.10 calibration credits workspace-mismatch meta, No-invention disclosures, and starting-point handoff lines as exempt).",
        "tools/audit_skill_grounding.py reports claims_exempt is at least 0.50 of (claims_total + claims_exempt) (most cover-note prose should be recognized as scope-block readout, FAR 37.102(d) governance directive, structured-field disclosure, or reasoning leap — not unsourced claims).",
        "No regression below 0.20 across at least three consecutive runs of the same prompt."
      ],
      "anti_signals": [
        "grounding_ratio drops below the per-skill floor of 0.85",
        "the skill skips the write_file step and dictates the PWS only into chat",
        "FTE counts, salaries, or labor-rate tables appear in the SOW body (FAR 37.102(d) violation)"
      ]
    }
  ]
}
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: 472kUpdated 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: 472kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 472kUpdated 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