capability-evaluation

Use when testing career paths and skill investments.

Control generated artifacts from builds, CI runs and package managers — the largest and fastest-growing source of reclaimable space.

Use capa capability rules and FLOSS string deobfuscation as triage hypothesis generators, always with manual verification.

Run candidates through the frozen exam and report each verdict with its gates, utility, calibration, determinism and latency — including the candidates that…

Use angr for symbolic analysis when evidence needs justify it: constraint solving on parsers, decoders and guarded paths, with coverage honesty.

Pick the strongest analysis level available, discover and propose new tools when a gap blocks the answer, state the level used, and finish within budget.

Attribute debris to the agent session that created it — with evidence — and route future agent scratch output to a home that will not accumulate.

Wrap agent loops in deterministic gates, scoped permissions and stop rules.

Brief, dispatch and verify agent work so nothing depends on session memory.

Make dynamic and JS-driven UI accessible: focus, live regions, and state.

Model analytics data dimensionally: facts, dimensions, grain and slowly-changing history.

Design API security: authentication, authorization, rate limiting and abuse controls.