SkillAgentSearch skills...

igce-builder-ffp

Trigger for: FFP IGCE, firm-fixed-price estimate, FFP cost model, proposed FFP rate validation, wrap-rate analysis, Agency BPA rate comparison, price-reasonableness memo, or fair-and-reasonable analysis.

Install / Use

npx skills add aiskillstore/marketplace --skill igce-builder-ffp

Installs into whichever agent you are using.

About this skill
📄

SKILL.md

Installable skill definition

Quality Score

89/100

Supported Platforms

Universal

Tags

Our assessment of igce-builder-ffp

igce-builder-ffp scores 89/100 on our quality scale, 1728th of 4,717 Development & Engineering skills we index (top 37%).

Its SKILL.md is 22 KB long, well organised into 25 sections with 2 code examples: a thorough specification that gives an agent plenty to work with.

It has 428 GitHub stars, a meaningful sign that others use it.

Substance
30/30
Structure
18/20
Description
15/15
Adoption
11/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated about 2 months ago, so igce-builder-ffp is actively maintained.
  • Our last check on 2026-10-31 found the source still online.
  • No license is declared. By default that means all rights are reserved: you can read it, but reusing or redistributing it is not clearly permitted. Ask the author before building on it commercially.
  • Its trust signals score 88/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.

AI review by kimi-k2.7-code on 2026-10-07. Automated pattern scan on 2026-10-07. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

igce-builder-ffp compared with similar skills

All 4 of these similar skills score higher than igce-builder-ffp; compare them before choosing.

SkillScoreStarsUpdatedFormat
igce-builder-ffp (this skill)by aiskillstore8942855d agoSKILL.md
taste-skillby Leonxlnx100118.3k10d agoCLAUDE.md
hyperframesby heygen-com10086.4ktodayCLAUDE.md
ai-job-searchby ai-job-search-dev10072.4ktodayCLAUDE.md
claude-howtoby luongnv8910047.4k4d agoCLAUDE.md

Frequently asked questions

How do I install igce-builder-ffp?
Run npx skills add aiskillstore/marketplace --skill igce-builder-ffp. The install tabs above show the steps for each supported agent.
Which AI agents does igce-builder-ffp work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is igce-builder-ffp safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It declares no license and scores 88/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is igce-builder-ffp still maintained?
The repository was last updated about 2 months ago, so igce-builder-ffp is actively maintained.

name: igce-builder-ffp description: > Trigger for: FFP IGCE, firm-fixed-price estimate, FFP cost model, proposed FFP rate validation, wrap-rate analysis, Agency BPA rate comparison, price-reasonableness memo, or fair-and-reasonable analysis. Build auditable Firm-Fixed-Price federal estimates using BLS OEWS wages, layered fringe/overhead/G&A/profit, GSA CALC+ positioning, and GSA Per Diem travel. Use for FFP-by-period, FFP-by-deliverable, SOW/PWS decomposition, implied-multiplier analysis, and fixed-price scenario comparisons. Do NOT use for Labor Hour, T&M, or cost-reimbursement IGCEs (use IGCE Builder LH/T&M or IGCE Builder CR). Do NOT use for grant budgets. Requires the bls-oews, gsa-calc, and gsa-perdiem MCP servers.

IGCE Builder: Firm-Fixed-Price (FFP)

Purpose and operating boundary

Build an auditable FFP Independent Government Cost Estimate from BLS wages, layered indirect rates, CALC+ positioning data, travel inputs, and the Contracting Officer's assumptions. Keep the BLS wage, aging adjustment, fringe, overhead, G&A, profit, and fixed-price calculations visible and formula driven.

Assemble data and format workbooks. Do not originate professional judgments reserved to the Contracting Officer:

  • Do not determine that a price or rate is fair and reasonable.
  • Do not invent premiums for clearance, SCIF, OCONUS, specialty labor, or other conditions that the available data does not quantify.
  • Do not draft a determination, negotiation position, responsibility finding, or signature-ready FAR memorandum unless the user supplies the rationale and conclusion. If supplied, reproduce that text verbatim and mark the output DRAFT.
  • Use neutral positioning language such as "at CALC+ P77" or "above P50 by 22%." Avoid "reasonable," "defensible," "acceptable," "competitive," and similar conclusions.
  • Workflow B produces positioning data only unless the user chooses the controlled memo-fill path and supplies the required determination text.

If a sentence concludes whether a number is right or wrong, stop and replace it with sourced data, arithmetic, and the decision left to the Contracting Officer.

Reference map

Read only the references needed for the active workflow:

Non-negotiable gates

These gates prevent documented silent wrong answers. Keep them active even when shortening or adapting the workflow.

  1. CALC+ query signature: Use the CALC+ /v3/api/ceilingrates/ endpoint with the keyword= parameter when a keyword query is required. Never use q=. The wrong parameter can silently return the full corpus. Preserve the discovery path aggregations.labor_category.buckets with each bucket's key and doc_count in the data-source reference.
  2. Cross-sheet hourly-rate index: In each 19-row Cost Buildup block, row 4 is Aged Annual Wage and row 5 is Direct Labor Rate (Hourly). Summary, scenario, and validation formulas that need hourly direct labor must reference 5 + (i-1)*19, never 4 + (i-1)*19. The wrong row produced a documented $16.9 billion result.
  3. Month-gap formula: Store BLS vintage and contract start as YYYY-MM text and compute months with VALUE(LEFT(...)) and VALUE(MID(...)). Do not use YEAR() on text and do not substitute DATEDIF for this tested pattern.
  4. Step 8.5 validation: Run the formula-structure audit, independent Python recomputation, and real spreadsheet-engine verification when available. Never call an openpyxl-only check recalculation or proof of formula execution.
  5. AI boundary: Never originate a fair-and-reasonable determination. Workflow B is data only unless the user supplies verbatim Option B rationale and determination text.
  6. BLS vintage: Treat May 2025 only as the current documented baseline. Call detect_latest_year at runtime and use its result. Never age wages from a stale hardcoded vintage.
  7. Rate-positioning bands: Report 0-15% above CALC+ P50 as the expected range, 15-40% as the FFP premium band, and anything above 40% with explicit stacked-factor arithmetic. These are positioning bands, not determinations.
  8. Staged questions: For SOW/PWS decomposition, complete Stage A decomposition confirmation before Stage B build parameters. The Stage A response must end immediately after the decomposition-confirmation question. Do not preview, list, or request any Stage B input in that response. End each response at its question and wait. Never self-approve either stage.
  9. Aging factor: Put the aging assumptions in named or clearly labeled assumption cells. Reference those cells from formulas. Never hardcode an aging multiplier into labor formulas or methodology prose.
  10. Credentialed API pacing: Serialize calls to credentialed federal APIs and leave at least three seconds between calls. Never parallelize keyed calls. Honor a longer server-provided retry interval, and stop with a rate-limit report instead of rapid retrying.

Pre-flight: capabilities and dependencies

Select the workflow first, then run this before its first MCP-dependent step. For Workflow B, emit the required boundary and wait for the user's option before making a tool call.

  1. Inspect the operations available in the current session. Match by advertised MCP server and operation name or by equivalent operation schema. Do not depend on a host-generated namespace or separator.
  2. Confirm the operation groups required by the active workflow:
    • Workflows A and A+ require bls-oews operations detect_latest_year, get_wage_data, and the SOC/metro lookups, plus gsa-calc operations suggest_contains, exact_search, keyword_search, igce_benchmark, and price_reasonableness_check.
    • Add gsa-perdiem operations estimate_travel_cost, lookup_city_perdiem, and get_mie_breakdown only when travel is in scope.
    • Workflow B requires the listed bls-oews and gsa-calc operations but not gsa-perdiem unless the user also requests travel analysis.
  3. Before a build calls a data source, confirm an .xlsx authoring capability plus Python 3.10 or later with openpyxl for the bundled validators. A real spreadsheet engine is preferred but optional when its absence is disclosed exactly as required in Step 8.5.
  4. Test only capabilities the active workflow will use. For a build, call detect_latest_year. If travel is in scope, test the Per Diem capability when it is first needed. For Workflow B, test BLS and CALC+ only after the user selects an option. Apply the credentialed API pacing gate to every test call. Do not expose keys or credentials.
  5. If an operation is unavailable, look for a semantically equivalent operation exposed by the same server. Do not replace the MCP with a hand-built public API call.
  6. If a required capability remains missing, stop and list it. State whether it appears uninstalled, unauthenticated, or unavailable in the current host.

Use this message when installation is missing:

The active workflow requires these MCP capabilities: [list]. Missing: [list]. Install and configure them in this client, restart or refresh the client, and try again.

Use this message when authentication is missing:

[server] is available, but its required API key is not configured or was rejected. Configure the provider key in the MCP server, restart or refresh the client, and try again.

Select a workflow

Workflow A: Full FFP IGCE build

Use when the user supplies structured labor and contract inputs. Execute Steps 1 through 9.

Workflow A+: SOW/PWS-driven FFP build

Use when the user supplies a SOW, PWS, unstructured requirement, or the approved staffing handoff from sow-pws-builder. For an unstructured requirement, execute Step 0 and obtain both staged confirmations when required. For an approved handoff, consume it as described below and do not repeat decomposition or Stage A.

Workflow B: FFP rate positioning

Use when the user asks whether proposed FFP rates are reasonable, asks to validate a proposal, or requests price-reasonableness analysis.

On the first Workflow B response, do not call tools or begin analysis. Emit this boundary and stop:

I can pull positioning data that shows where each proposed rate sits against CALC+ ceiling rates and BLS market wages. I cannot draft a price reasonableness memo, write a "fair and reasonable" determination, or recommend negotiation positions. Those are Contracting Officer decisions under FAR 15.404-1, not skill outputs.

Tell me which you want:

Option A: Positioning data only. I produce a table with each proposed rate, CALC+ P25/P50/P75/P90 and sample size, plus a BLS metro burdened equivalent. I provide no verdict or recommendation.

Option B: Memo template fill. You provide your rationale and determination. I reproduce your text verbatim in a DRAFT memo and place the benchmark tables underneath it. I do not originate conclusions or negotiation positions.

Which option?

Proceed only after the user selects Option A or supplies both Option B rationale and determination text. For Option A, return neutral benchmark tables and stop. For Option B, reproduce user-supplied rationale and determination verbatim, mark the memo DRAFT, and use placeholders for the Contracting Officer and agency.

Collect inputs

Collect missing information in the fewest useful stages. Use the host's structured question tool when it is available. Otherwise present numbered choices in chat and accept a number, label, or free-text answer.

Required for a build:

  • Labor categories or priceable task areas
  • Performance location or metro
  • FTE or other staffing basis
  • Productive hours per person, default 1,880
  • Period of performance and pricing structure
  • Contract start month
  • Contract vehicle or indirect-rate basis

Optional with disclosed defaults:

  • Fringe, overhead, G&A, profit, and escalation rates
  • Travel destinations, frequency, nights, travelers, months, and origin
  • NAICS and PSC
  • Partial-period months
  • Shift-coverage requirement
  • Deliverable weights or per-LCAT allocation matrix

Do not guess a required discipline, location, staffing basis, vehicle, or pricing structure.

Consume an approved SOW/PWS handoff

Treat a table labeled STAFFING HANDOFF TABLE and identified for the IGCE Builder as user-reviewed input, regardless of the separator punctuation in its heading. Do not decompose the requirement again.

  1. Confirm that the declared contract type is FFP. For a hybrid, accept only the FFP CLINs and leave LH/T&M and CR CLINs to their respective skills.
  2. Preserve each approved Labor Category, SOC Code, FTE, Phase, Hours/Yr, Notes, derivation, and user override. Do not silently remap or resize staffing.
  3. Use the companion CLIN handoff when present to populate the period or deliverable pricing structure.
  4. Reconcile any contradiction between the handoff, SOW/PWS, and current user instruction in a short table and wait for the user to choose which value controls.
  5. Ask for all missing pricing inputs in one Stage B response: performance location, contract start, vehicle or indirect-rate basis, period mapping, travel or ODC assumptions, and any required allocation choice. Use a structured question tool or one numbered list, end at the question, and wait. Ask sequential one-field questions only when an answer changes the available c

Truncated for display — read the full file on GitHub.

Related Skills

View on GitHub
GitHub Stars428
CategoryDevelopment
Updated1mo ago
Forks45

Languages

Python

Trust signals

88/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 medium