Chat-only: build order analysis for modframe_wallstreet vs modframe_wallstreet_dc#1
Closed
Chat-only: build order analysis for modframe_wallstreet vs modframe_wallstreet_dc#1
Conversation
…r linear ordering
…uthoring
Adds the full scaffolding layer needed for a weaker model to draft, source,
and review topic modules against enforceable contracts.
New scripts:
- scripts/validate_completeness.py — schema-driven validator; checks required
headings, placeholder detection, Status field controlled vocabulary;
supports --warn-only / --stats-only / --status filters
- scripts/generate_queue.py — generates prioritized work queue from live topic
state; text/csv/json output; --next N flag for task assignment
New schemas (docs/schemas/):
- outline_schema.json — required headings, placeholder strings, word count
targets, claim type labels; loaded by validator (edit JSON, not Python)
- readme_schema.json — valid Status values, required fields, recommended tags
New prompt packs (docs/prompts/):
- draft_module.md — initial draft with embedded style rules, {{variable}} slots,
word count targets, post-draft checklist
- citation_pass.md — source enrichment with tier hierarchy and format rules
- neutrality_pass.md — structured framing review with 5 flag categories and
required output format
- final_polish.md — prose quality pass with change log requirement
GitHub process scaffolding:
- .github/PULL_REQUEST_TEMPLATE.md — checklist tied to lifecycle states
- .github/ISSUE_TEMPLATE/new_module.yml — structured form for topic requests
- .github/ISSUE_TEMPLATE/correction.yml — structured form for factual corrections
Updated docs:
- docs/GOVERNANCE.md — 6-state lifecycle (empty→scaffolded→draft→sourced→
reviewed→published) with transition rules and anti-capture section
- docs/MODULE_TEMPLATE.md — lifecycle table, precise outline template with
intent-annotated placeholders, claim label rules, production workflow
- .github/workflows/docs-quality.yml — adds completeness check step (warn-only)
- docs/examples/README.md — golden examples criteria and structure
Design notes: work queue is generated on demand (not stored as 308 task.json
files); schema rules are externalized so non-technical contributors can update
them without touching Python; --warn-only allows gradual CI gate rollout.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…ring Single document a content-producing LLM reads at session start to understand its role, constraints, and step-by-step workflow. Covers: - What the library is and is not - What the model must never do - Banned language list with reasoning - Claim type label rules (Observed/Inferred/Hypothesis) with gate conditions - Citation format and source tier hierarchy - Full lifecycle (empty→sourced) with per-step checklists - Quick-reference table pointing to all supporting tools and docs Distinct from docs/prompts/ (which are task-specific prompts to send to a model) — this is the orientation/system-prompt-level document. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Modules moved from empty stubs to Status: sourced. Each contains: substantive prose, [Observed]/[Inferred]/[Hypothesis] claim labels, ≥7 real primary sources with full citation format, real case studies in Episode outline Part 3, and ≥4 concrete process steps. 001_Campaign_Finance_Architecture - Covers FECA architecture, two-track system (disclosed limited vs. unlimited independent), super PAC / 501(c)(4) dark money pass-through mechanism - Sources: FECA (52 U.S.C.), Buckley, Citizens United, McCutcheon, OpenSecrets, FEC enforcement stats, CRS RL30965 - Case study: Crossroads GPS / American Crossroads in 2012 cycle 002_Primaries_and_Candidate_Gatekeeping - Covers primary types (closed/open/top-two), ballot access requirements, money primary dynamics, electorate-nominee mismatch mechanism - Sources: CA Dem Party v. Jones, NCSL, Ballotpedia, Hall (2019), Brookings, DNC/RNC delegate selection rules - Case study: Eric Cantor 2014 primary loss 003_Gerrymandering_Systems - Covers post-census redistricting authority, Rucho federal court withdrawal, efficiency gap metric, commission vs. legislative process comparison - Sources: Rucho, VRA § 10301, Allen v. Milligan, Stephanopoulos/McGhee (2015), NCSL, Princeton Gerrymandering Project, Dave's Redistricting App - Case study: REDMAP 2010 strategy and Wisconsin efficiency gap Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
From post-run quality review of 001-003:
Fix 1 — Claim label syntax
outline_schema.json: replace vague claim_labels block with
claim_label_formats; adds canonical forms for all four variants
(bare [Observed], inline [Observed — source: SHORT_REF], [Inferred],
[Hypothesis]) with usage note explaining when each applies.
Fix 2 — Validator: claim label count
validate_completeness.py now counts [Observed/Inferred/Hypothesis]
occurrences and fails if fewer than min_claim_labels (5). Catches
modules with correct headings but unlabeled prose.
Fix 3 — Validator: source entry count + actors word count
Parses Suggested sources section, counts real (non-placeholder) bullet
entries; checks actors section word count. Both thresholds read from
outline_schema.json (sources_min_entries: 3, actors_min_words: 80).
extract_section() handles variable heading depth correctly.
Fix 4 — Topic context step in draft prompt
draft_module.md: adds {{TOPIC_CONTEXT}} variable and instructions for
writing a 4-6 sentence knowledge brief before drafting. CONTENT_AGENT.md:
inserts Step 3 (write topic brief) before Step 4 (draft); marks it
mandatory with rationale. Addresses cold-start weakness where a weak model
has only the topic slug as context before generating full module prose.
Fix 5 — Golden examples
Copied 001-003 to docs/examples/ with updated README listing all three
examples, their source counts, and usage guidance (calibration, prompt
grounding, reviewer reference). Previous README was a placeholder; now
live with current content.
All three sourced modules continue to pass validate_completeness.py
after new checks are applied.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…CV modules Also fixes MD001 markdownlint rule (heading increment) across all module files by disabling it in .markdownlint.json — the outline.md format intentionally uses H1 title → bold Summary/Mechanism fields → H3 sections, skipping H2 by design. Applies to all 308 topic files retroactively. 004_Voter_Access_Mechanisms - Covers state-federal patchwork, NVRA/HAVA/VRA floor requirements, and the structural shift caused by Shelby County v. Holder (2013) eliminating Section 5 preclearance — converting the check from pre-enforcement to post-enactment - Actors: state legislatures, election officials, EAC, DOJ Civil Rights, courts - Failure modes: preclearance elimination gap, photo ID demographic cost asymmetry, voter roll purge errors, polling place consolidation lag - Sources: NVRA (52 U.S.C.), Shelby County, Crawford, Brnovich, EAC EAVS, MIT Cost of Voting Index, Hajnal et al. (2017) - Case study: Texas strict photo ID law — blocked under Section 5, enacted immediately post-Shelby County, found discriminatory intent in 2017 but governed multiple elections during litigation (Veasey v. Abbott) 005_Ballot_Initiative_Systems - Covers direct vs. indirect, statutory vs. constitutional variants; professional signature-gathering industry; ballot title framing power; legislative rollback - Actors: sponsors, paid petitioning firms, legislature, Secretary of State, courts, national campaign funders - Failure modes: Meyer v. Grant commodification effect, legislative implementation narrowing (Florida Amendment 4 / SB 7066), single-subject inconsistency, out-of-state funding capture - Sources: NCSL, Meyer v. Grant, Gerber Populist Paradox, IRI, Ballotpedia, California Elections Code, Jones v. Governor of Florida - Case study: Florida Amendment 4 (2018) — 64% voter approval for returning citizens' voting rights; legislature conditioned restoration on fee payment 006_Ranked_Choice_Voting_Mechanics - Covers elimination rounds, ballot exhaustion, tabulation delay, equipment certification bottleneck, legislative preemption patterns - Actors: state/local governments, election administrators, equipment vendors/ certifiers, candidates (incentive structure), voters, FairVote (advocacy) - Failure modes: ballot exhaustion, tabulation delay, equipment certification barrier, state preemption of local adoption - Sources: Maine RCV Act, Alaska Ballot Measure 2, Alaska DoE 2022 results, NCSL, FairVote (with advocacy-source caveat), Burnett/Kogan (2015), Donovan et al. (2016) - Case study: Alaska 2022 special election — Palin/Peltola/Begich; first federal RCV general election; Begich eliminated, his voters' second choices elected Peltola despite Palin leading first-round count Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…rden CONTENT_AGENT Module 007 — Electoral College Allocation Rules: - Full outline: winner-take-all vs. district allocation, elector process, ECRA 2022 certification reforms, faithless electors post-Chiafalo - Case study: Nebraska LB 764 (2024) — allocation rules as partisan lever - 9 primary sources: Art. II/XII, ECRA 2022 P.L. 117-328, CRS R47031, Chiafalo, Bush v. Gore, NARA, NCSL elector methods and faithless elector laws - validate_completeness: 7/7 sourced modules clean CONTENT_AGENT.md additions to "What you must never do": - Explicit ban on creating new scripts, tools, schemas, or infrastructure - Ban on persistent files for intermediate steps (no TOPIC_CONTEXT.txt etc.) - Ban on building replacement tooling when commands fail — ask instead - Ban on stopping to ask which option mid-workflow without cause Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
007 — Electoral College Allocation Rules: - Pew Research: specify section URL rather than homepage - Larcinese et al.: note institutional access / Google Scholar path - Nebraska LB 764: add legislature session number and searchable URL 008 — State Certification Processes: - Summary expanded from 2 to 4 sentences (adds Safe Harbor deadline, ECRA 2022 conclusive certificate requirement) - State courts actor: replace "state case law examples" placeholder with Wood v. Raffensperger and Pearson v. Kemp real citations - Competing certificates: add Wayne County canvassing board episode (Nov 2020) with Michigan Board of State Canvassers source - Partisan pressure: replace unanchored [Inferred] with documented Raffensperger call (Jan 2, 2021), sourced to GA SoS transcript and Senate Judiciary report - Evidence section: fix bare [Observed] with real case citations - NARA URL: /electoral-college/about → /electoral-college/ascertainment - Brennan Center replaced with EAC Post-Election Audits guidance (tier-1 federal) - Georgia 2020 case study: expanded from 2 lines to 12 lines with vote margin, three-review sequence, court citations, Raffensperger call, congressional certification date - Added 4 new source entries: Wood v. Raffensperger, Raffensperger call transcript, Senate Judiciary "Subverting Justice" report, EAC audit guidance Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…enriched citations, expanded content - Back-patched 35 asymmetric cross-references across 13 modules - Normalized 69 actor names to 47 canonical entries (15 now multi-module) - Added 1-2 statutes and 1 case to each batch module (031-042) - Expanded batch module outlines from ~630 to ~745 avg words - Rebuilt indexes (actors, legal_citations, related_modules)
…43–065) - 23 outlines authored with full sourced content, claim labels, and citations - 23 READMEs updated: status sourced, actors, statutes, cases, cross-refs - All 65 modules pass metadata and completeness validators (65/65 clean) - Indexes rebuilt: 147 actors, 147 legal citations, 253 related edges
- Back-patched 11 asymmetric cross-references across 7 READMEs - Normalized 58 actor names to 43 canonical entries (19 renames across 18 files) - Indexes rebuilt: 131 actors (was 147), 264 related edges (was 253) - All 65 modules pass both validators (65/65 clean)
Section 01 (001-017): - Replace placeholder citations with topically relevant statutes and cases (2+ each) - Normalize actor YAML indentation from spaces to tabs - Fix asymmetric cross-ref: add 013, 017 to module 001 related modules - Remove self-reference from 001 related modules Section 02 (018-030): - Add 2nd statute to 021-030 (10 modules) - Add 2nd case to 018-030 (13 modules) - Add 5th actor bullet to 019-032 outlines (14 modules) - Add 4th failure-mode bullet to 019-030 outlines (12 modules) - Module 020 word count: 647 → 691 (from added actor bullet) Indexes rebuilt: 192 legal citations (+45), 265 related edges (+1) Validators: 65/65 clean on metadata and completeness
- 23 outlines covering lobbying, cooling-off periods, K Street, revolving door, FARA, think tanks, PACs, bundlers, shadow lobbying, astroturf, trade assocs, 501c4s, judicial nominations, state preemption, tech/pharma/energy lobbying, contractor feedback loops, ethics waivers, and transparency gaps - 23 READMEs with tab-indented YAML actors, 2+ statutes, 2+ cases each - Validators: 88/88 clean on both metadata and completeness - Indexes rebuilt: 193 actors, 264 legal citations, 358 related edges
Co-authored-by: sethc5 <206268795+sethc5@users.noreply.github.com>
Copilot
AI
changed the title
[WIP] Review repository and provide feedback
Remove duplicate section 04 that caused topic ID validation failures
Feb 19, 2026
Copilot
AI
changed the title
Remove duplicate section 04 that caused topic ID validation failures
Chat-only: applications of Modframe once fully built out
Feb 19, 2026
Copilot
AI
changed the title
Chat-only: applications of Modframe once fully built out
Chat-only: legislative and public communication applications of Modframe for engaged legislators
Feb 19, 2026
Copilot
AI
changed the title
Chat-only: legislative and public communication applications of Modframe for engaged legislators
Chat-only: Polis + Modframe civic leverage app concept
Feb 19, 2026
Copilot
AI
changed the title
Chat-only: Polis + Modframe civic leverage app concept
Remove duplicate section 04 stub directory causing topic ID validation failures
Feb 19, 2026
Copilot
AI
changed the title
Remove duplicate section 04 stub directory causing topic ID validation failures
Chat-only: build order analysis for modframe_wallstreet vs modframe_wallstreet_dc
Feb 19, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Strategic discussion on sequencing the two stub repos. No code changes.
Conclusion
Build
modframe_wallstreetfirst —modframe_wallstreet_dcis explicitly a bridge layer and cannot be meaningfully constructed until both upstream libraries have sourced content.Key findings
wallstreet_dcrequiresRelated modulesIDs, index bundles, and crosswalk data from both upstream repos by designmodframe_wallstreetis closer to ready than it looks: Sections 04 (revolving door/regulatory capture), 08 (money & industry influence), and 19 (financial/economic policy levers) inmodframealready cover the DC side of financial power —wallstreetonly needs to add the market-structure side (SEC/CFTC/OCC/FRB territory)modframewith minimal adaptationstaleness_queue.pywill be more powerful in the Wall Street variant than in the DC variantRecommended sequence
modframe→modframe_wallstreetmodframecontentmodframe_wallstreet_dcbridge modules only after both upstream sections are sourced — start with mechanisms already fully resolved in both (e.g.,modframe072 Financial Sector Regulatory Capture ↔ wallstreet Section 4 SEC Enforcement)💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.