ciregenz eae31e744f [eric] browser: fold page text into the auto-attached post-action state (OSW_FOLD_TEXT, default off) = after any action the model has BOTH clickable elements AND readable text, so it stops double-perceiving (measured: 52% of tool calls are perception, ~half redundant list+GetText pairs, the biggest untouched turn-cost)

2026-07-08 17:08:16 -07:00

..

__init__.py

[eric] tree: group browser sub-agent modules under agents/browser

2026-05-23 13:06:55 -07:00

browser_agent.py

[eric] browser: fold page text into the auto-attached post-action state (OSW_FOLD_TEXT, default off) = after any action the model has BOTH clickable elements AND readable text, so it stops double-perceiving (measured: 52% of tool calls are perception, ~half redundant list+GetText pairs, the biggest untouched turn-cost)

2026-07-08 17:08:16 -07:00

browser_batch_replay.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_extract.py

[eric] browser: leading-_ -> p_/P_/public for module functions+constants+state across browser subsystem; promote cross-file symbols public (detect_loop, BROWSER_HISTORY, TRUSTED, etc.), align analyze-browser-metrics.py + tests

2026-06-23 19:37:40 -07:00

browser_fast_path.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_fast_read.py

[eric] browser: flip speed levers default-ON (hop/prestage/trim/evict, kill switch =0); conftest pins them off for the mocked suite

2026-07-06 14:12:34 -07:00

browser_history.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_loop.py

[eric] browser: verified-action executor = resolve-late resolver + act/verify/re-aim step loop + BrowserActVerified tool (model emits 2-4 dependent click/fill steps in ONE turn, code resolves each by name against the live page, verifies the specific effect, re-aims once; irreversible targets refused in code, solo-send rule unchanged); 16 unit + 1 loop-interception test

2026-07-07 22:42:39 -07:00

browser_meta_playbook.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_metrics.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_plan_dispatch.py

[eric] browser: planner maps ordinals to the Nth matching row (live fire declined 'the 4th story's comments' because it wasn't told it may count)

2026-07-07 23:26:43 -07:00

browser_playbook.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_prestage.py

[eric] browser: prestage aux nudge for person-messaging tasks (stage through the profile Message surface, never a feed detour; r245 staged to /feed/)

2026-07-07 02:41:27 -07:00

browser_save.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_schema.py

[eric] browser: reframe BrowserBatch schema copy (was 'your standard way to ACT', steering the model away from the verified tool; now points dependent named click/fill sequences at BrowserActVerified)

2026-07-07 23:04:30 -07:00

browser_self_audit.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_send_script.py

[eric] browser: verified-action executor foundation = generic site-agnostic effect verification (did the action produce the SPECIFIC expected effect: url_changed/appeared/gone/filled/cleared), zero per-site code; wired into the send-script's receipt (proves it on the real flow, behavior-identical to the old inline check via an equivalence test) so it's the executor's verification core, not dead scaffolding

2026-07-07 18:00:35 -07:00

browser_skills.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

browser_validator.py

[eric] browser: leading-_ -> p_/P_/public for module functions+constants+state across browser subsystem; promote cross-file symbols public (detect_loop, BROWSER_HISTORY, TRUSTED, etc.), align analyze-browser-metrics.py + tests

2026-06-23 19:37:40 -07:00

browser_verified_action.py

[eric] browser: verified-action executor = resolve-late resolver + act/verify/re-aim step loop + BrowserActVerified tool (model emits 2-4 dependent click/fill steps in ONE turn, code resolves each by name against the live page, verifies the specific effect, re-aims once; irreversible targets refused in code, solo-send rule unchanged); 16 unit + 1 loop-interception test

2026-07-07 22:42:39 -07:00

browser_verified_step.py

[eric] browser: verified-action executor = resolve-late resolver + act/verify/re-aim step loop + BrowserActVerified tool (model emits 2-4 dependent click/fill steps in ONE turn, code resolves each by name against the live page, verifies the specific effect, re-aims once; irreversible targets refused in code, solo-send rule unchanged); 16 unit + 1 loop-interception test

2026-07-07 22:42:39 -07:00

browser_wait.py

[eric] backend: one-line every comment + collapse 3+ blank runs across apps/ (AST-verified identical, -lines); CLAUDE.md one-line-comment rule

2026-06-25 02:29:04 -07:00

seed_for.py

[eric] backend: rename single-export modules to match their export (naming convention sweep)

2026-06-24 01:30:04 -07:00