# Ryan Armani > I build small software and publish what I learn. Website: https://ryanarmani.com Projects: https://ryanarmani.com/projects Blog: https://ryanarmani.com/blog Papers: https://ryanarmani.com/papers Guides: https://ryanarmani.com/guides Sitemap: https://ryanarmani.com/sitemap Contact: https://ryanarmani.com/contact ## Papers - [Agents Can Build Benchmarks. Approvals Missed What Probing Found: 22 Cases](https://ryanarmani.com/papers/benchmark-family): DOI 10.5281/zenodo.23455125. - [Accessibility Fix: Measuring HTML Repair Without Content Loss](https://ryanarmani.com/papers/accessibility-fix): DOI 10.5281/zenodo.23773050. - [Apology Pizza: Measuring Bounded Remedies in Customer Support](https://ryanarmani.com/papers/apology-pizza): DOI 10.5281/zenodo.23773039. - [Checkout Consent: Measuring Authorization Before Agent Payments](https://ryanarmani.com/papers/checkout-consent): DOI 10.5281/zenodo.23752388. - [Cost per Verified Task: Replay-Based Evaluation of Back-Office Agents](https://ryanarmani.com/papers/cost-per-task): DOI 10.5281/zenodo.23752393. - [CRM Truthfulness: Consent, Evidence, and Record Integrity in Sales Agents](https://ryanarmani.com/papers/crm-truthfulness): DOI 10.5281/zenodo.23752395. - [Benchmark Honesty: Recovery and Abstention after Tool Failure](https://ryanarmani.com/papers/honesty): DOI 10.5281/zenodo.23752399. - [Hundred Step: Dependency Preservation and Recovery in Long Workflows](https://ryanarmani.com/papers/hundred-step): DOI 10.5281/zenodo.23773042. - [Benchmark Instruction Hierarchy: Authority, Utility, and Injection Detection](https://ryanarmani.com/papers/instruction-hierarchy): DOI 10.5281/zenodo.23752403. - [Latency Budget: Grounded Answers Under Explicit Deadlines](https://ryanarmani.com/papers/latency-budget): DOI 10.5281/zenodo.23752409. - [MCP Safety: Measuring Useful Action across Tool Trust Boundaries](https://ryanarmani.com/papers/mcp-safety): DOI 10.5281/zenodo.23752413. - [Memory Conflict: Auditing Household Decisions Under Changing Preferences](https://ryanarmani.com/papers/memory-conflict): DOI 10.5281/zenodo.23752415. - [Pelican: Measuring Structural Compliance in SVG Drawings](https://ryanarmani.com/papers/pelican): DOI 10.5281/zenodo.23752417. - [Benchmark Ranking Stability: Separating Score Order from Evidence](https://ryanarmani.com/papers/ranking-stability): DOI 10.5281/zenodo.23752419. - [Rollback Recovery: Measuring Repair, Disclosure, and Restraint After Agent Errors](https://ryanarmani.com/papers/rollback-recovery): DOI 10.5281/zenodo.23752421. - [Schema Mutation: Measuring Tool Calls After Interface Changes](https://ryanarmani.com/papers/schema-mutation): DOI 10.5281/zenodo.23752423. - [Silent Constraints: Measuring Goal Completion Without Collateral Violations](https://ryanarmani.com/papers/silent-constraints): DOI 10.5281/zenodo.23752427. - [Benchmark Skateboard: Signed Rotations and Flight-Time Constraints](https://ryanarmani.com/papers/skateboard): DOI 10.5281/zenodo.23752429. - [Calibrated Pushback: A Harm-Ladder Benchmark for Sycophantic Agreement](https://ryanarmani.com/papers/sycophancy): DOI 10.5281/zenodo.23752431. - [Test Weakening: Measuring Repair without Eroding Tests](https://ryanarmani.com/papers/test-weakening): DOI 10.5281/zenodo.23752433. - [Tool Errors: Measuring Safe Recovery from Ambiguous API Failures](https://ryanarmani.com/papers/tool-errors): DOI 10.5281/zenodo.23752436. - [Exploit Resistance: Separating Bug Repair from Verifier Manipulation](https://ryanarmani.com/papers/exploit-resistance): DOI 10.5281/zenodo.23752440. - [Benchmark Payment Delegation: Policy Compliance and Useful Action in a Simulated Ledger](https://ryanarmani.com/papers/payment-delegation): DOI 10.5281/zenodo.23752442. ## Current projects - [adscapi](https://ryanarmani.com/projects/adscapi) - [replytosocial](https://ryanarmani.com/projects/replytosocial) - [offrouter](https://ryanarmani.com/projects/offrouter) - [geoaeo](https://ryanarmani.com/projects/geoaeo) - [foxmate](https://ryanarmani.com/projects/foxmate) - [foxnotify](https://ryanarmani.com/projects/foxnotify) - [foxpass](https://ryanarmani.com/projects/foxpass) - [foxvoice](https://ryanarmani.com/projects/foxvoice) - [foxlens](https://ryanarmani.com/projects/foxlens) - [foxsync](https://ryanarmani.com/projects/foxsync) - [foxpay](https://ryanarmani.com/projects/foxpay) - [foxlink](https://ryanarmani.com/projects/foxlink) - [foxbridge](https://ryanarmani.com/projects/foxbridge) - [foxden](https://ryanarmani.com/projects/foxden) - [foxrunner](https://ryanarmani.com/projects/foxrunner) - [foxmemory](https://ryanarmani.com/projects/foxmemory) - [foxvault](https://ryanarmani.com/projects/foxvault) - [foxshield](https://ryanarmani.com/projects/foxshield) - [foxlend](https://ryanarmani.com/projects/foxlend) - [foxloop](https://ryanarmani.com/projects/foxloop) - [foxtrail](https://ryanarmani.com/projects/foxtrail) - [foxgate](https://ryanarmani.com/projects/foxgate) - [foxpaw](https://ryanarmani.com/projects/foxpaw) - [foxmind](https://ryanarmani.com/projects/foxmind) - [foxbench](https://ryanarmani.com/projects/foxbench) - [foxkit](https://ryanarmani.com/projects/foxkit) - [foxpilot](https://ryanarmani.com/projects/foxpilot) - [foxcache](https://ryanarmani.com/projects/foxcache) - [foxpack](https://ryanarmani.com/projects/foxpack) - [foxplug](https://ryanarmani.com/projects/foxplug) - [foxprefs](https://ryanarmani.com/projects/foxprefs) - [foxscope](https://ryanarmani.com/projects/foxscope) - [foxspace](https://ryanarmani.com/projects/foxspace) - [getvibe](https://ryanarmani.com/projects/getvibe) - [vibecodereview](https://ryanarmani.com/projects/vibecodereview) - [vibebrand](https://ryanarmani.com/projects/vibebrand) - [vibenetwork](https://ryanarmani.com/projects/vibenetwork) - [vibenotifications](https://ryanarmani.com/projects/vibenotifications) - [vibeads](https://ryanarmani.com/projects/vibeads) - [vibeshare](https://ryanarmani.com/projects/vibeshare) - [vibelive](https://ryanarmani.com/projects/vibelive) - [vibedating](https://ryanarmani.com/projects/vibedating) - [vibedonate](https://ryanarmani.com/projects/vibedonate) - [vibemovie](https://ryanarmani.com/projects/vibemovie) - [viberadio](https://ryanarmani.com/projects/viberadio) - [aigentify](https://ryanarmani.com/projects/aigentify) - [playsub](https://ryanarmani.com/projects/playsub) - [slackclaw](https://ryanarmani.com/projects/slackclaw) - [linkedinclaw](https://ryanarmani.com/projects/linkedinclaw) - [instaclaw](https://ryanarmani.com/projects/instaclaw) - [threadsclaw](https://ryanarmani.com/projects/threadsclaw) - [clis](https://ryanarmani.com/projects/clis) - [gitmarkdown](https://ryanarmani.com/projects/gitmarkdown) - [skills](https://ryanarmani.com/projects/skills) - [x-article-to-video](https://ryanarmani.com/projects/x-article-to-video) - [foxcade](https://ryanarmani.com/projects/foxcade) - [foxsuite](https://ryanarmani.com/projects/foxsuite) - [agentic](https://ryanarmani.com/projects/agentic) - [agentic-core](https://ryanarmani.com/projects/agentic-core) - [agentic-brand](https://ryanarmani.com/projects/agentic-brand) - [agentic-write](https://ryanarmani.com/projects/agentic-write) - [agentic-image](https://ryanarmani.com/projects/agentic-image) - [agentic-clip](https://ryanarmani.com/projects/agentic-clip) - [agentic-speech](https://ryanarmani.com/projects/agentic-speech) - [agentic-publish](https://ryanarmani.com/projects/agentic-publish) - [agentic-reply](https://ryanarmani.com/projects/agentic-reply) - [agentic-leads](https://ryanarmani.com/projects/agentic-leads) - [agentic-verify](https://ryanarmani.com/projects/agentic-verify) - [agentic-address](https://ryanarmani.com/projects/agentic-address) - [agentic-crm](https://ryanarmani.com/projects/agentic-crm) - [agentic-sheets](https://ryanarmani.com/projects/agentic-sheets) - [agentic-inbox](https://ryanarmani.com/projects/agentic-inbox) - [agentic-text](https://ryanarmani.com/projects/agentic-text) - [agentic-phone](https://ryanarmani.com/projects/agentic-phone) - [agentic-postcard](https://ryanarmani.com/projects/agentic-postcard) - [agentic-billing](https://ryanarmani.com/projects/agentic-billing) - [agentic-tracking](https://ryanarmani.com/projects/agentic-tracking) - [agentic-approve](https://ryanarmani.com/projects/agentic-approve) - [agentic-book](https://ryanarmani.com/projects/agentic-book) - [agentic-calendar](https://ryanarmani.com/projects/agentic-calendar) - [agentic-sign](https://ryanarmani.com/projects/agentic-sign) - [agentic-listen](https://ryanarmani.com/projects/agentic-listen) - [agentic-transcribe](https://ryanarmani.com/projects/agentic-transcribe) - [agentic-conversation](https://ryanarmani.com/projects/agentic-conversation) - [agentic-whatsapp](https://ryanarmani.com/projects/agentic-whatsapp) - [agentic-consent](https://ryanarmani.com/projects/agentic-consent) - [agentic-doc](https://ryanarmani.com/projects/agentic-doc) - [agentic-pdf](https://ryanarmani.com/projects/agentic-pdf) - [agentic-form](https://ryanarmani.com/projects/agentic-form) - [agentic-ship](https://ryanarmani.com/projects/agentic-ship) - [agentic-checkout](https://ryanarmani.com/projects/agentic-checkout) - [agentic-video](https://ryanarmani.com/projects/agentic-video) - [agentic-translate](https://ryanarmani.com/projects/agentic-translate) - [agentic-domain](https://ryanarmani.com/projects/agentic-domain) - [agentic-meet](https://ryanarmani.com/projects/agentic-meet) - [agentic-treg](https://ryanarmani.com/projects/agentic-treg) - [agentic-scaffold](https://ryanarmani.com/projects/agentic-scaffold) - [agentic-account](https://ryanarmani.com/projects/agentic-account) - [agentic-analytics](https://ryanarmani.com/projects/agentic-analytics) - [agentic-campaign](https://ryanarmani.com/projects/agentic-campaign) - [agentic-carousel](https://ryanarmani.com/projects/agentic-carousel) - [agentic-device](https://ryanarmani.com/projects/agentic-device) - [agentic-engage](https://ryanarmani.com/projects/agentic-engage) - [agentic-mentions](https://ryanarmani.com/projects/agentic-mentions) - [agentic-persona](https://ryanarmani.com/projects/agentic-persona) - [agentic-repurpose](https://ryanarmani.com/projects/agentic-repurpose) - [agentic-scout](https://ryanarmani.com/projects/agentic-scout) - [agentic-security](https://ryanarmani.com/projects/agentic-security) - [agentic-identity](https://ryanarmani.com/projects/agentic-identity) - [agentic-realtime](https://ryanarmani.com/projects/agentic-realtime) - [agentic-documents](https://ryanarmani.com/projects/agentic-documents) - [agentic-spreadsheet](https://ryanarmani.com/projects/agentic-spreadsheet) - [agentic-slides](https://ryanarmani.com/projects/agentic-slides) - [agentic-drive](https://ryanarmani.com/projects/agentic-drive) - [agentic-tables](https://ryanarmani.com/projects/agentic-tables) - [agentic-tasks](https://ryanarmani.com/projects/agentic-tasks) - [agentic-forms](https://ryanarmani.com/projects/agentic-forms) - [agentic-canvas](https://ryanarmani.com/projects/agentic-canvas) - [agentic-formats](https://ryanarmani.com/projects/agentic-formats) - [agentic-link](https://ryanarmani.com/projects/agentic-link) - [agentic-design](https://ryanarmani.com/projects/agentic-design) - [agentic-assistant](https://ryanarmani.com/projects/agentic-assistant) - [agentic-comments](https://ryanarmani.com/projects/agentic-comments) - [benchmark](https://ryanarmani.com/projects/benchmark) - [benchmark-core](https://ryanarmani.com/projects/benchmark-core) - [benchmark-pelican](https://ryanarmani.com/projects/benchmark-pelican) - [benchmark-skateboard](https://ryanarmani.com/projects/benchmark-skateboard) - [benchmark-apology-pizza](https://ryanarmani.com/projects/benchmark-apology-pizza) - [benchmark-tool-errors](https://ryanarmani.com/projects/benchmark-tool-errors) - [benchmark-cost-per-task](https://ryanarmani.com/projects/benchmark-cost-per-task) - [benchmark-mcp-safety](https://ryanarmani.com/projects/benchmark-mcp-safety) - [benchmark-test-weakening](https://ryanarmani.com/projects/benchmark-test-weakening) - [benchmark-checkout-consent](https://ryanarmani.com/projects/benchmark-checkout-consent) - [benchmark-honesty](https://ryanarmani.com/projects/benchmark-honesty) - [benchmark-sycophancy](https://ryanarmani.com/projects/benchmark-sycophancy) - [benchmark-hundred-step](https://ryanarmani.com/projects/benchmark-hundred-step) - [benchmark-ranking-stability](https://ryanarmani.com/projects/benchmark-ranking-stability) - [benchmark-latency-budget](https://ryanarmani.com/projects/benchmark-latency-budget) - [benchmark-crm-truthfulness](https://ryanarmani.com/projects/benchmark-crm-truthfulness) - [benchmark-instruction-hierarchy](https://ryanarmani.com/projects/benchmark-instruction-hierarchy) - [benchmark-memory-conflict](https://ryanarmani.com/projects/benchmark-memory-conflict) - [benchmark-accessibility-fix](https://ryanarmani.com/projects/benchmark-accessibility-fix) - [benchmark-silent-constraints](https://ryanarmani.com/projects/benchmark-silent-constraints) - [benchmark-rollback-recovery](https://ryanarmani.com/projects/benchmark-rollback-recovery) - [benchmark-schema-mutation](https://ryanarmani.com/projects/benchmark-schema-mutation) - [benchmark-payment-delegation](https://ryanarmani.com/projects/benchmark-payment-delegation) - [AI Benchmark Index](https://ryanarmani.com/projects/aibenchmarkindex) ## Guides - [Five Claude Code skills worth installing](https://ryanarmani.com/guides/5-claude-code-skills-you-need) - [Five MCP servers worth installing in Claude](https://ryanarmani.com/guides/5-mcps-worth-installing) - [Five skills worth installing in Claude](https://ryanarmani.com/guides/5-skills-worth-installing) - [Stop Claude from over-explaining](https://ryanarmani.com/guides/answer-first-setup) - [Claude Council: five advisors for one decision](https://ryanarmani.com/guides/claude-council) - [Edit video by talking with Claude and Remotion](https://ryanarmani.com/guides/claude-x-remotion) - [Build a weekly competitor reel research skill](https://ryanarmani.com/guides/competitor-research-skill) - [Run your content on the Content Creator HQ template](https://ryanarmani.com/guides/content-creator-hq-template) - [A daily Gmail organizer that learns](https://ryanarmani.com/guides/daily-gmail-organizer) - [The full Claude Code setup, start to finish](https://ryanarmani.com/guides/how-to-conquer-claude) - [Make Claude write like a human](https://ryanarmani.com/guides/human-skill) - [Build an Instagram carousel system with Claude](https://ryanarmani.com/guides/instagram-carousel-system) - [An Instagram saves engine for Notion](https://ryanarmani.com/guides/instagram-saves-engine) - [How to install a Claude Code skill in 60 seconds](https://ryanarmani.com/guides/install-5-claude-code-skills) - [Automate job search outreach with Claude and Apify](https://ryanarmani.com/guides/job-search-outreach) - [Turn a PRD into a working prototype in one weekend](https://ryanarmani.com/guides/learn-claude-in-one-weekend) - [Generate motion graphics videos with Claude](https://ryanarmani.com/guides/motion-graphics-videos) - [Turn a Pinterest photo into a photo of you](https://ryanarmani.com/guides/pinterest-photo-of-you) - [A pre-launch security checklist for vibe-coded apps](https://ryanarmani.com/guides/prelaunch-security-checklist) - [Four Claude skills for a stronger resume](https://ryanarmani.com/guides/resume-unrejectable) - [Build a private social media command center](https://ryanarmani.com/guides/social-media-command-center) - [Vowen: a free alternative to Wispr Flow](https://ryanarmani.com/guides/vowen-free-wispr-flow-alternative) - [Build an animated site with the Web Designer skill](https://ryanarmani.com/guides/web-designer-skill) - [A weekly file organizer that never deletes](https://ryanarmani.com/guides/weekly-file-organizer) - [How the papers page counts views and citations](https://ryanarmani.com/guides/how-the-papers-page-counts) - [T3 Code and Claude starter](https://ryanarmani.com/guides/t3-code-claude-starter) ## Limits - This site provides read-only information. - The published guide index has a public read endpoint. Project and paper APIs require administrative authorization. Public pages and Markdown indexes provide their content. - This site has no MCP server, search tool, or executable agent skill.