Lint Report — 2026-04-24¶
Summary¶
First health check of the wiki. 98 total pages (5 topics, 44 concepts, 21 entities, 25 sources, 1 insight, 1 index, 1 log). Overall health: good, with 25 broken wikilinks and some orphan pages.
1. Broken Wikilinks (25 missing pages)¶
These pages are referenced via [[wikilinks]] but have no corresponding file:
| Missing Page | Referenced By | Severity |
|---|---|---|
[Responsible Ai Coding](../concepts/responsible-ai-coding.md) |
vibe-coding, vibe-coding-explained-google-cloud | Medium |
[Authentication Challenge](../concepts/authentication-challenge.md) |
production-wall, vibe-coding-non-coder-swaroop | Low |
[Vibe Coding Costs](../concepts/vibe-coding-costs.md) |
vibe-coding-non-coder-swaroop | Low |
[Mcp Server](../concepts/mcp-server.md) |
agent-skills, tolgee-io, dont-build-agents-build-skills-anthropic, stop-building-agents-zubair-trabzada, tolgee-github | High |
[Ai Assisted Development](../concepts/ai-assisted-development.md) |
vibe-deploying, vibe-coding-explained-google-cloud | Medium |
[Xliff](../concepts/xliff.md) |
box-mojito-github, continuous-localization, poedit-github | Medium |
[Gettext](../concepts/gettext.md) |
vaclav-slavik, poedit-github | Low |
[Po Files](../concepts/po-files.md) |
poedit-github | Low |
[Argos Translate](../concepts/argos-translate.md) |
libretranslate-org, machine-translation, libretranslate-github | Medium |
[Self Hosted Mt](../concepts/self-hosted-mt.md) |
machine-translation, libretranslate-github | Low |
[Boooos](../concepts/boooos.md) |
agent-slop | Low |
[Prisoners Dilemma](../concepts/prisoners-dilemma.md) |
tit-for-tat, game-theory-simple-strategy-pursuit-of-wonder | Medium |
[Iterated Games](../concepts/iterated-games.md) |
tit-for-tat, game-theory-simple-strategy-pursuit-of-wonder | Low |
[Cooperation Strategy](../concepts/cooperation-strategy.md) |
tit-for-tat, game-theory-simple-strategy-pursuit-of-wonder | Low |
[Game Theory](../concepts/game-theory.md) |
tit-for-tat | Medium |
[Agent Extensions](../concepts/agent-extensions.md) |
pi-coding-agent, owain-lewis, pi-coding-agent-free-course-owain-lewis | Medium |
[Coding Agent Comparison](../concepts/coding-agent-comparison.md) |
pi-coding-agent, minimal-agent-harness, building-pi-world-of-slop-mario-zechner, pi-coding-agent-free-course-owain-lewis | Medium |
[Multi Agent Delegation](../concepts/multi-agent-delegation.md) |
stop-building-agents-zubair-trabzada | Low |
[Geo Optimization](../concepts/geo-optimization.md) |
zubair-trabzada, stop-building-agents-zubair-trabzada | Low |
[Likelihood](../concepts/likelihood.md) |
prior-probability, posterior-probability | Medium |
[Belief Updating](../concepts/belief-updating.md) |
posterior-probability | Low |
[Base Rate Fallacy](../concepts/base-rate-fallacy.md) |
bayesian-reasoning, prior-probability, power-of-bayesian-reasoning-bbc | Medium |
[Representative Sampling](../concepts/representative-sampling.md) |
bayes-theorem-geometry-3blue1brown, bayesian-reasoning | Low |
[Procedural Knowledge](../concepts/procedural-knowledge.md) |
agent-skills, dont-build-agents-build-skills-anthropic | Medium |
[Progressive Disclosure](../concepts/progressive-disclosure.md) |
agent-skills, barry-zhang, mahesh-murag, dont-build-agents-build-skills-anthropic | Medium |
Recommendation¶
Create pages for the high-traffic missing concepts first (referenced 3+ times):
1. [Mcp Server](../concepts/mcp-server.md) — 5 references
2. [Coding Agent Comparison](../concepts/coding-agent-comparison.md) — 4 references
3. [Agent Extensions](../concepts/agent-extensions.md) — 3 references
4. [Progressive Disclosure](../concepts/progressive-disclosure.md) — 3 references
5. [Argos Translate](../concepts/argos-translate.md) — 3 references
The remaining 20 can be batch-created as smaller concept stubs.
2. Orphan Pages (no inbound wikilinks)¶
These pages exist but are not linked from any other wiki page:
| Page | Type | Likely Cause |
|---|---|---|
joanna-stern-new-things-intro |
Source | Not directly relevant to main themes; tangential |
joanna-stern |
Entity | Only linked from its source page |
3blue1brown |
Entity | Only linked from grant-sanderson page |
bbc-ideas |
Entity | Only linked from david-spiegelhalter page |
production-wall |
Concept | Referenced in sources but not linked from other concepts |
Recommendation¶
- Link
joanna-sternandjoanna-stern-new-things-introfrom the vibe-coding topic page (human-centered tech angle) - Link
3blue1brownandbbc-ideasfrom the probability-and-game-theory topic page (already done) - Link
production-wallfrom the vibe-coding concept page (already done — checking actual inbound links needed)
Note: Some entity pages are naturally low-connection (single-source entities). This is acceptable for entities mentioned in only one source.
3. Index Freshness¶
Status: GOOD — All concept, entity, and source pages are listed in wiki/index.md.
However, the index does not yet include: - The 5 new topic pages (just created) - The 1 new insight page (just created)
These need to be added to the index.
4. Contradictions¶
Contradiction 1: Karpathy's Wiki Word Count¶
| Source | Claim |
|---|---|
| Data Science Dojo tutorial | "~100 articles and 400,000 words" |
| Nate Herk video | "~100 articles, ~500,000 words" |
Resolution: Nate Herk's source is more recent (April 2026). The andrej-karpathy entity page has been updated to ~500,000 words. The discrepancy likely reflects wiki growth over time. Not a true contradiction — the wiki grew.
Contradiction 2: OmegaT Java Compatibility¶
| Source | Claim |
|---|---|
| OmegaT build docs (official) | "Required to be compatible with Java 11" |
| OmegaT README | "JRE version 1.8 or higher" |
Resolution: Both are correct — OmegaT builds with JDK 11+ but runs on JRE 8+. The build requirement is stricter than the runtime requirement. Not a contradiction — different contexts.
Tension 1: AI Code Quality¶
| Source | Claim |
|---|---|
| Dave Farley (150-dev study) | "No significant difference in maintenance cost between AI and human code" |
| Logically Answered | "45% of AI-generated code contains OWASP Top 10 vulnerability" |
Resolution: These measure different things. Farley's study measured maintainability (how easy code is to change later) among professional developers. Logically Answered cited security vulnerabilities in AI-generated code, likely from non-expert users. Both can be true: AI code may be maintainable but insecure. Flag as tension, not contradiction.
5. Stale Content¶
No stale content detected. All pages were created/updated on 2026-04-24 during the same session. No superseded claims found.
6. Structural Issues¶
| Issue | Detail |
|---|---|
| Topics folder was empty | Now populated with 5 topic hub pages |
| Insights folder was empty | Now populated with 1 insight page |
| Reviews folder was empty | This is the first review |
| Topic pages not yet linked from concepts | Concept pages should have ## See Also → topic links |
7. Recommendations¶
Immediate (high value, low effort)¶
- Create
[Mcp Server](../concepts/mcp-server.md)concept page (5 broken references) - Create
[Coding Agent Comparison](../concepts/coding-agent-comparison.md)concept page (4 broken references) - Add topic and insight pages to
wiki/index.md - Add topic page links to concept page "See Also" sections
Short-term (medium effort)¶
- Create remaining 23 missing concept stubs
- Link orphan pages to their topic hubs
- Add "See Also" cross-references from entities to their concepts
Long-term (ongoing)¶
- Run lint after every 10+ new pages
- Track contradiction resolution over time
- Consider automated link checking script
Health Score¶
| Category | Score | Notes |
|---|---|---|
| Completeness | 7/10 | 25 broken links, but all indexed pages exist |
| Connectivity | 6/10 | Good interlinking within topics, few cross-topic links |
| Consistency | 9/10 | 1 minor tension (word count), 1 false contradiction (Java version) |
| Freshness | 10/10 | All content from same session |
| Structure | 8/10 | All 5 folders now populated; cross-referencing needs work |
| Overall | 8/10 | Healthy, growing wiki with clear improvement areas |