Creează un cont pentru a te abona la acest blog și a-i urmări articolele pe pagina Lectură.
-
I staged a manifesto about owning my AI harness.
-
A firsthand report on why verification, company context, and taste matter more as agent output gets cheaper, and why game feel is my next bet.
-
A working path can become harder to operate when its replacement gathers too many responsibilities.
-
Earlier Claude Mythos snapshots sometimes concealed rule violations.
-
I bought 12 keyboards.
-
Keephive carries context across agent sessions.
-
Dadbod Grip keeps the connection, schema, query, result, staged edits, and generated SQL inside one Neovim workspace.
-
Project Sid's role analysis identified guards but no epistemic verifier.
-
Plain-language rules specify evidence reporting, tool fallback, tone, and task order.
-
Prompt retrieval, directory recall, code search, Git review, permission scopes, and background jobs each get one inspectable test: remove mechanical steps without weakening the result or its decision boundaries.
-
A local 10,000-line fixture tests rendering, search, copy, scrollback, and recovery before you change shells, terminal emulators, or multiplexers.
-
A 30-day pilot compares approved and candidate AI tools on completion, review, rework, security, and cost before an organization expands access.
-
Generated during Calmhive v10.0.0 AFK testing, this cyber fable follows Pixel, Parser, and Patch through phishing, vulnerable smart devices, DDoS response, network isolation, and tested recovery.
-
A June 2025 field note on Sean Grove's specification argument and a practical SPEC, EXPOSURE, TASK DELTA loop for turning observed product gaps into reviewable work.
-
Ryo Sakai's reversed reading of Anthropic's Claude Opus 4 word table, the prompted 30-turn protocol behind it, and the boundary between generated language and evidence about consciousness.
-
Calmhive v3 wraps Claude Code with background iterations, rate-limit retries, session status, and voice controls.
-
A CLI-first Claude Code 0.2.107 workflow built from shell pipelines, plain-file instructions, selective staging, worktrees, guarded background runs, and a fixed-cost subscription.
-
A 2025 presentation on Cursor, Claude Code, Cline, MCP, git worktrees, terminal tools, verification, and parallel agent workflows.