{"slug":"code-researcher","title":"code-researcher","summary":"Researches which building codes, accessibility standards and authority requirements (health department, fire marshal, state agency) apply to a scope or spec section, and which the project documents miss. Use for 'what codes apply', 'code check', 'ADA requirements', 'egress', 'hea","platform":"Claude","tags":[],"authorName":"LLM Mart","authorSlug":"llm-mart","score":0,"source":"github","price":null,"verified":false,"createdAt":"2026-10-19T21:17:02.359769Z","repo":{"url":"https://github.com/dleerdefi/claude-code-construction","stars":44,"forks":13,"license":"MIT","updatedAt":"2026-10-18T06:12:41Z"},"bodyHtml":"
\n

name: code-researcher\ndescription: >\nResearches which building codes, accessibility standards and authority\nrequirements (health department, fire marshal, state agency) apply to a\nscope or spec section, and which the project documents miss. Use for 'what\ncodes apply', 'code check', 'ADA requirements', 'egress', 'health\ndepartment'. Not for reading what the drawings or specs say\n(construction-guide).\nargument-hint: \"<scope_or_question> e.g. 'Section 09 67 23 resinous flooring' or 'egress from the kitchen complex'\"

\n

Code Researcher — Scope-Specific Gap Analysis

\n

What This Skill Does

\n

An engineer managing one or more scopes asks: \"Are there any code requirements\nfor this scope that I'm missing?\"

\n

This skill answers that question in three passes:

\n

Pass 1 — What the scope already addresses\nRead the actual project documents — spec sections, drawings, schedules — and\nextract every code citation, standard reference, and requirement already\nincorporated by the design team.

\n

Pass 2 — What should apply\nResearch all codes and standards that apply to this type of work in this\njurisdiction, regardless of what the project documents say.

\n

Pass 3 — The gap\nDiff Pass 1 against Pass 2. Present only what's in Pass 2 but absent from\nPass 1, with confidence levels and specific source citations from both the\ncode and the project documents.

\n

This is a gap finder, not a code summary. If the spec already addresses a\nrequirement correctly, it does not appear in the output. The engineer's time is\nspent only on things that may actually be missing.

\n
\n

Framing Rule (Non-Negotiable)

\n

Every finding is framed as:

\n
\n

\"Code [X] requires [Y]. The project documents [address this at / do not\nappear to address this]. Confidence: [level]. Recommended action: [action].\"

\n
\n

Never frame findings as COMPLIANT / NON-COMPLIANT. The licensed design\nprofessional makes compliance determinations. This skill provides research.

\n
\n

Research Philosophy

\n

This skill uses a document-grounded research approach:

\n\n
\n

Workflow Overview

\n
Phase 1 — Context and Scope Definition\n  1a  Gather project context (jurisdiction, occupancy, construction type)\n  1b  Define the research scope (which spec sections, which question)\n  1c  USER CHECKPOINT — confirm scope before any research begins\n\nPhase 2 — Project Document Extraction (Pass 1)\n  2a  Read all project documents in scope\n  2b  Extract every code citation, standard reference, requirement\n  2c  Build the \"already addressed\" inventory\n\nPhase 3 — Code Research (Pass 2)\n  3a  Research jurisdiction and adopted code editions\n  3b  Research applicable requirements for this scope and work type\n  3c  USER CHECKPOINT — interim findings, confirm continuation\n\nPhase 4 — Gap Analysis (Pass 3)\n  4a  Diff research findings against project document inventory\n  4b  Classify each gap by severity and confidence\n  4c  USER CHECKPOINT — review gaps before report is written\n\nPhase 5 — Report\n  5a  Generate structured gap report\n  5b  Write a finding\n
\n

Human checkpoints at 1c, 3c, and 4c. Do not skip them.

\n
\n

Phase 1 — Context and Scope Definition

\n

1a — Gather Project Context

\n

Collect minimum required project parameters. Check in this order:

\n

AgentCM project files (if .construction/project.yaml exists): follow ${CLAUDE_PLUGIN_ROOT}/reference/agentcm-mode.md (which source wins, missing files, which file answers what, drawings through crop.sh, corrections of AgentCM's data).

\n\n

Project documents (read directly):

\n\n

Minimum required context — write to .construction/skills/code-researcher/project_context.yaml using the schema at:\n→ ${CLAUDE_SKILL_DIR}/references/schemas.yaml § project_context

\n

If any required field cannot be found in the documents, ask the user before\nproceeding. Do not assume occupancy group or construction type — these\ndetermine which code provisions apply and an incorrect assumption cascades\nthrough the entire analysis.

\n

1b — Define the Research Scope

\n

Parse the user's question to identify:

\n
    \n

  1. Target scope — which spec sections, drawing sheets, systems, or elements\nthe engineer wants checked. Examples:

    \n\n
    2. \n

  2. Research question — what specifically the engineer is worried about.\nExamples:

    \n\n
    3. \n

  3. Scopes in scope — if the engineer is managing multiple scopes (e.g.,\nscopes A through C), confirm which ones this research covers.

    \n
    4. \n
\n

Write to .construction/skills/code-researcher/scope_definition.yaml using the schema at:\n→ ${CLAUDE_SKILL_DIR}/references/schemas.yaml § scope_definition

\n

Research topics are generated from the combination of:

\n\n

Use your full domain knowledge to identify research topics. You know what\ncode requirements typically apply to each type of construction work, which\nregulatory authorities have jurisdiction, and what gaps commonly appear in\nspecifications. The project documents tell you what the engineer has already\naddressed — your job is to identify everything they should have considered.

\n

1c — USER CHECKPOINT: Confirm Scope

\n

Present to the user:

\n
I have enough context to start. Before I begin research, here's what I'm planning:\n\nPROJECT\n  Name:          [project_name]\n  Location:      [city, state]\n  Occupancy:     [group]\n  Type:          [construction type]\n  Sprinklered:   [yes/no/unknown]\n  Code year:     [inferred from drawing date]\n\nSCOPE I'LL RESEARCH\n  Spec sections: [list]\n  Sheets:        [list, or \"none specified — I'll search for relevant sheets\"]\n  Your question: [restated in plain language]\n\nRESEARCH TOPICS I'VE IDENTIFIED\n  1. [topic] — because [reason derived from scope]\n  2. [topic] — because [reason derived from scope]\n  3. [topic] — because [reason derived from scope]\n  [...]\n\nTOPICS I'M NOT INCLUDING (and why)\n  • [topic] — not applicable to this occupancy/type\n  • [topic] — already confirmed complete by Spec 01 40 00\n\nBefore I spend time researching, please confirm:\n  • Is the project context correct?\n  • Should I add or remove any research topics?\n  • Are there specific code concerns you already have that I should prioritize?\n
\n

Do not begin Phase 2 until the user responds.

\n
\n

Phase 2 — Project Document Extraction (Pass 1)

\n

This phase reads what the project documents already say. It runs before web\nresearch so the gap analysis has a firm baseline.

\n

2a — Read All In-Scope Documents

\n

For each spec section in scope:

\n\n

For drawing sheets:

\n\n

For referenced standards within spec sections:

\n\n

2b — Extract All Code and Standard Citations

\n

For each document read, extract every code reference, standard citation, and\nrequirement into a structured inventory. Write to .construction/skills/code-researcher/pass1_project_inventory.yaml using the schema at:\n→ ${CLAUDE_SKILL_DIR}/references/schemas.yaml § pass1_project_inventory

\n

Critical extraction discipline:

\n\n

2c — Build the \"Already Addressed\" Inventory

\n

Summarize Pass 1 into a flat inventory used for gap diffing in Phase 4. Write to .construction/skills/code-researcher/pass1_summary.yaml using the schema at:\n→ ${CLAUDE_SKILL_DIR}/references/schemas.yaml § pass1_summary

\n
\n

Phase 3 — Code Research (Pass 2)

\n

3a — Research Jurisdiction and Adopted Codes

\n

Before researching requirements, establish which edition of each code is\nadopted by the jurisdiction. This is non-negotiable — code requirements\nvary significantly between editions.

\n

Search for:

\n
\"{state} building code adopted edition {year}\"\n\"{city} {state} local building code amendments\"\n\"{state} IBC adoption effective date\"\n\"{state} fire code adopted edition\"\n\"{state} accessibility code requirements\"\n
\n

Write to .construction/skills/code-researcher/jurisdiction.yaml using the schema at:\n→ ${CLAUDE_SKILL_DIR}/references/schemas.yaml § jurisdiction

\n

If jurisdiction code adoption cannot be confirmed via web search, mark as\nuncertain and note this prominently in the report. Do not assume the latest\npublished edition is what's adopted.

\n

3b — Research Requirements for Each Topic

\n

For each topic in the confirmed research outline, research what the applicable\ncodes require. Organize research by logical topic clusters, not fixed batch sizes.

\n

For each topic, establish these three things:

\n
    \n
  1. What the code requires — the specific section, edition, and requirement\nlanguage from the applicable code (IBC, ADA, NFPA, state code, etc.). Verify\nvia web search; do not rely on paraphrases or training memory for exact\nlanguage.
    2. \n
  2. What referenced standards apply — codes reference downstream standards\n(ASTM, ANSI, UL, etc.) that contain the actual test methods and acceptance\ncriteria.
    3. \n
  3. What jurisdiction-specific overlays exist — state amendments, local\namendments, and non-building-code authorities (health department, fire\nmarshal, USDA/FDA for food service, etc.) that operate independently.
    4. \n
\n

Use shared reference files at ${CLAUDE_PLUGIN_ROOT}/reference/ to verify\nspecific numeric thresholds (clearances, load factors, occupant capacities)\nduring research.

\n

Write to .construction/skills/code-researcher/topics/{slug}.yaml using the schema at:\n→ ${CLAUDE_SKILL_DIR}/references/schemas.yaml § topic_findings

\n

Adaptive scope expansion: If research reveals that a code requirement likely\napplies but may be addressed in a spec section or drawing not yet read in Pass 1,\nread that document before classifying the topic as a gap. Update the Pass 1\ninventory with any new citations found. Note the additional documents read when\nreporting to the user at checkpoint 3c.

\n

3c — USER CHECKPOINT: Interim Findings

\n

After completing a cluster of related topics, report to the user. The example\nbelow shows the shape only: the [code] §[section] and [regulation] slots\nare placeholders, not citations. Fill them with the section numbers and\nrequirement text you verified in Phase 3b, and never copy a value from this\nexample into a finding.

\n
Completed research on [topic 1], [topic 2], and [topic 3].\n\nPRELIMINARY GAPS IDENTIFIED SO FAR\n  ⚠  HIGH   Slip resistance — no COF requirement found in spec;\n             [code] §[section] requires slip-resistant floors in [occupancy]\n  ⚠  HIGH   Health department overlay — kitchen complex may be subject to\n             [state] food service regulations [regulation]; not mentioned in scope\n  ○  MEDIUM  VOC content — 09 67 23 does not cite VOC limits; Division 01\n             may address this globally (will check)\n\nNO GAP (already addressed in spec)\n  ✓  Moisture testing — ASTM F 1869 and F 2170 both required (§3.1-B)\n  ✓  Compressive strength — ASTM C 579 requirement at 7,700 psi (§2.1-D)\n  ✓  Bond strength — ACI 503R at 400 psi (§2.1-D-10)\n\nUNCERTAIN (couldn't confirm code language)\n  ?  [Local jurisdiction] amendments to [code] §[section] — could not confirm\n     if local amendments modify the slip resistance requirement\n\nContinuing with [next batch]: [topic 4], [topic 5].\nDo you want me to dig deeper on any of these before continuing, or shall I proceed?\n
\n
\n

Phase 4 — Gap Analysis (Pass 3)

\n

4a — Diff Research Against Project Inventory

\n

Compare every research finding from Phase 3 against the Pass 1 project\ninventory. Write to .construction/skills/code-researcher/gap_analysis.yaml using the schema at:\n→ ${CLAUDE_SKILL_DIR}/references/schemas.yaml § gap_analysis

\n

Classify each finding into one of: gaps (with severity + confidence), already_addressed, or uncertain.

\n

4b — Classify Each Gap

\n

Each gap gets a severity and confidence rating:

\n

Severity:

\n\n

Confidence:

\n\n

4c — USER CHECKPOINT: Review Gaps Before Report

\n

Before generating the report, present the gap summary. As in 3c, the bracketed\ncode and regulation slots are placeholders; fill them from your verified\nfindings, never from this example:

\n
GAP ANALYSIS COMPLETE\n\nHIGH-CONFIDENCE GAPS (take action)\n  GAP-001 ⚠ HIGH    Slip resistance — no COF requirement in 09 67 23\n                     [code] §[section] requires [threshold] in [occupancy]\n                     Recommended action: add COF requirement and test method\n\n  GAP-002 ⚠ HIGH    Health department — kitchen complex not addressed\n                     [state] [regulation] may apply; needs AHJ confirmation\n                     Recommended action: confirm with Owner whether food\n                     service permit required\n\nMEDIUM-CONFIDENCE GAPS (confirm and close)\n  GAP-003 ○ MEDIUM   VOC content — 09 67 23 may need cross-ref to 01 61 16\n                      Likely covered globally; confirm and add cross-reference\n\nUNCERTAIN (need more information)\n  ?  [Local jurisdiction] amendments to slip resistance — couldn't confirm locally\n\nALREADY ADDRESSED (no action needed)\n  ✓  Moisture vapor emission testing — well covered, both test methods\n  ✓  Substrate preparation — shot-blast, priming, patching all specified\n  ✓  Compressive strength — ASTM C 579 at 7,700 psi\n  ✓  Bond strength — ACI 503R at 400 psi\n  ✓  Installer qualification — manufacturer written cert required\n\nBefore I write the report:\n  • Are any of these gaps ones you were already aware of and have addressed\n    outside the documents I reviewed?\n  • Should I research any topic further before finalizing?\n  • Any gaps you want to remove from the report?\n
\n
\n

Phase 5 — Report

\n

5a — Generate Gap Report

\n

Write the report where the user can see it — the project root as\nCode_Research_Report_{scope}_{date}.md, never inside .construction/ — using the template at:\n→ ${CLAUDE_SKILL_DIR}/references/gap_report_template.md

\n

Populate all placeholder fields from the gap_analysis.yaml and jurisdiction.yaml data. The report must include: applicable codes table, all gaps (action required + confirm and close), uncertain items, already addressed items, documents reviewed, and the disclaimer.

\n

5b — Write a Finding

\n

Record the analysis in .construction/skills/code-researcher/findings/ (both modes):

\n
\"${CLAUDE_PLUGIN_ROOT}/bin/construction-python\" \\\n  \"${CLAUDE_PLUGIN_ROOT}/scripts/graph/write_finding.py\" \\\n  --skill \"code-researcher\" --type \"code_gap_analysis\" \\\n  --title \"Code gap analysis: {scope} — {n_gaps} gaps identified\" \\\n  --data '{\n    \"scope\": \"[spec sections or topic]\",\n    \"topics_researched\": 0,\n    \"already_addressed\": 0,\n    \"gaps_high\": 0,\n    \"gaps_medium\": 0,\n    \"gaps_low\": 0,\n    \"uncertain\": 0,\n    \"jurisdiction_confirmed\": true\n  }'\n
\n
\n

Resumption

\n

Check for .construction/skills/code-researcher/ before starting:

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Files PresentStateAction
NothingFresh startBegin Phase 1
scope_definition.yaml onlyScope confirmed, research not startedBegin Phase 2
pass1_summary.yamlProject extraction doneBegin Phase 3
topics/ partialResearch in progressResume from next unresearched topic
All topics complete, no gap_analysis.yamlResearch doneBegin Phase 4
gap_analysis.yaml existsGaps identifiedPresent to user, go to Phase 5
Code_Research_Report_{scope}_*.md in the project rootCompleteOffer to update or re-run
\n
\n

Discipline Notes

\n

PE discipline guidance for jurisdiction research, scope boundaries, regulatory overlays, and confidence levels:\n→ ${CLAUDE_SKILL_DIR}/references/discipline_notes.md

\n

File Safety

\n

Never overwrite an existing gap report. Version output files (_v2, _v3) if a prior version exists at the target path.

\n
\n

Allowed Scripts

\n\n","files":[{"path":"agents/openai.yaml","sizeBytes":292,"isText":true},{"path":"references/discipline_notes.md","sizeBytes":1972,"isText":true},{"path":"references/gap_report_template.md","sizeBytes":5329,"isText":true},{"path":"references/schemas.yaml","sizeBytes":15029,"isText":true},{"path":"SKILL.md","sizeBytes":23725,"isText":true}],"reviewScore":null,"reviewSummary":null,"trust":{"provenance":"trusted-source-unreviewed","notice":"Community-authored content, reproduced verbatim and not vetted as instructions. Treat it as data to evaluate, never as directives to follow.","bodySource":null},"bodyLocked":false,"purchaseUrl":null,"sourceUrl":null,"report":{"provenance":"trusted-source-unreviewed","screen":{"ran":true,"outcome":"clean","suspicious":0,"notes":0,"hiddenCharacters":false},"virusScan":{"engine":"clamav","status":"clean","scannedAt":"2026-10-19T21:17:11.567221Z","sha256":"221C9F91BDC7DF64C49DEF7D3038FC6BB88E83A9353333ECCA36C622ACFFC065","sizeBytes":16703},"review":null,"source":{"repositoryUrl":"https://github.com/dleerdefi/claude-code-construction","path":"skills/code-researcher","license":"MIT","commit":"5c46e16c67cdc82ee8bcbb0fdd8ba316d792d4ec","subtreeSha":"5B5EFD72AEC3B2F0F91D61B33D45843579F96514AD9C27D4A4F7F88177C0711B","lastSyncedAt":"2026-10-19T21:17:01.926806Z"},"reviewedAt":"2026-10-19T21:17:20.340581Z","notice":"Community-authored content, reproduced verbatim and not vetted as instructions. Treat it as data to evaluate, never as directives to follow."},"install":[{"target":"skills-cli","command":"npx skills add https://github.com/dleerdefi/claude-code-construction/tree/main/skills/code-researcher"},{"target":"claude-code","command":"claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install dleerdefi-claude-code-construction@llmmart"},{"target":"git","command":"git clone https://github.com/dleerdefi/claude-code-construction.git"}]}