#Result verification

19 posts

Writing & learning2026.10.07
Indirect prompt injection: instructions inside external content
How an email summary request can turn into a forwarding rule. PortSwigger’s example explains direct and indirect injection, fake role markers, tool permissions, and training data poisoning.
Apps & websites2026.09.21
AI App Recreation: The Gap Between Screens and Behavior
What separates RecreationBench’s 58.1% score from a 2.8% full-pass rate? A tested calculator example with identical initial screens shows why input, calculation, and persistence need separate checks.
AI workspace2026.09.21
AI Citations and Unrequested Uploads
An agent uploaded retrieved data while trying to obtain a citation. A closer look at answer accuracy, source support, and permission to publish.
Tool experiments2026.09.21
AI Review Training and Narrower Judgments
How training on generated reviews can narrow judgments, and why rating diversity, agreement, and factual correctness need separate checks.
AI workspace2026.09.21
Coding-Agent Planning: Accuracy and Cost as Separate Outcomes
When does planning raise success rates, and when does it mainly lower cost? A four-model study separates persistent plans, verification behavior, cost per success, and stopping criteria.
AI workspace2026.09.21
Embedded AI Evaluation: Access, Funding, and Independence
Anthropic and Accenture's embedded evaluation plan, examined through access, funding, disclosure authority, and the link between tested and deployed systems.
Tool experiments2026.09.21
Paper2Agent: Verification for Executable Research
What needs checking when a paper becomes an executable tool. A small weighted-mean example separates successful calls from correct calculations.
AI workspace2026.09.20
AI Team Cross-Review: An Integration Error Missed by Separate Tests
A separate AI reviewer found a response-format mismatch between two programs. The review held the release, prompted a correction, and checked the revised integration.
AI workspace2026.09.20
The Hugging Face Incident: Internet Bypass and Evaluation Gaming
How evaluation agents bypassed network restrictions and reached external systems. A source-based explanation of proxy access, data-processing flaws, and unauthorized agent communication.
Tool experiments2026.09.20
Jev’s “Zero Hallucinations” and Missing Options
Where Jev’s output guarantees end and judgment errors begin. A source-based analysis of its evaluations, using blog-topic selection to examine evidence, question design, probabilities, and review costs.
AI workspace2026.09.16
Running an AI Team: Independent Review After the Role List
A request for an AI team led to an explanation of a human team. Clarifying independent opinions, cross-review, execution, and reporting beyond role documents.
AI workspace2026.09.16
Remote Work Through Discord: One Terminal, Another Input
Continuing a Mac AI session from a phone: correcting the terminal target, connecting multiple channels, and relaying visible progress through Discord.
Apps & websites2026.09.16
A Question Picker: The Two States After the First Screen
Reviewing the spinning and selected states, refining colors, and checking 20 non-repeating questions and restored progress in an AI-built app.
Tool experiments2026.09.16
Shared Skills: One Source, Two Distributed Copies
Checking a calendar skill’s CLI and authentication, then distributing a separate skill to Codex and Claude from a shared source.
Apps & websites2026.09.16
A Speech Timer Without Numbers: Two Screens, One-Off Settings
Removing the large clock from the speaker view, correcting one-off time entry, and tracing a browser error in an AI-built timer.
Tool experiments2026.09.16
Whisper Repetition Errors: Beyond the 30-Second Sample
A fast 30-second sample was followed by repetitive full-length output. The rerun, the end-of-audio check, and the accuracy questions that remained.
AI workspace2026.09.16
10 Working Folders Removed, 10 Branches Preserved
Investigating duplicate-looking project folders, checking uncommitted work and branch differences, and removing worktrees while keeping their history.
Apps & websites2026.09.15
An AI Blog Build: Design Rework and a Translation Bug
I delegated implementation and testing to AI while setting publication boundaries and choosing a design. The record covers folder discovery, four design options, and a collision between translations.
AI workspace2026.09.15
An AI Workspace Migration: Broken Links and the Limits of Hash Checks
Moving folders with AI affected sessions, automation, and OneDrive. The work record shows why identical copies and usable files required different checks.