# AGENTS.md — Build guide for Browse AI ## Project scope Save a reviewed selector recipe for one catalogue table, preview extracted fields, schedule refreshes and show field-level changes. Catalogue verdict: kinda. The visible web data extraction loop is buildable, but a credible replacement needs more than the first screen. Browse AI earns its keep through connectors, auth, reliability, so expect a weekend or multi-day build and a narrower personal scope. Use the implementation prompt below to define the deliverable. Complete each phase's acceptance checks before extending the scope. ## Working agreement - Inspect the repository and its existing instructions before choosing paths, dependencies or commands. Keep one coherent stack and explain changes to the proposed architecture. - Plan a vertical slice that accepts a real input and produces the useful output described below. Persist only the state the prompt calls for; respect memory-only and upstream-managed workflows. Use fixtures only when they are clearly labelled. - After scaffolding, document the actual install, development, check and build commands in README and keep them synchronized with the package or project manifest. Do not report commands as successful unless they ran. - Work in small steps. At handoff, list implemented flows, checks actually performed, remaining blockers, and any credentials or provider setup the owner must supply. - Do not publish, spend money, contact customers, delete source data or run irreversible migrations without the project owner's authorization. ## Prerequisites - Runtime and tools: Python, HTTP parsing, SQLite and optional Playwright for explicitly permitted rendered pages. - Before starting: A small authorized URL set, reviewed selectors or field schema and a user-approved crawl frequency; no proxy-evasion machinery. ## Stack and architecture - Python, HTTP parsing, SQLite and optional Playwright for explicitly permitted rendered pages - Data design: Store RecipeVersion, PageSnapshot, RowIdentity and Observation; a stable product key separates a price change from a reordered table. Empty extraction is an error until reviewed. - Setup: A small authorized URL set, reviewed selectors or field schema and a user-approved crawl frequency; no proxy-evasion machinery ## Security and data integrity - Bound hosts, redirects, page count, body size and timeout. Reject private-network destinations throughout redirects, respect access restrictions and backoff, and distinguish extraction failure from a missing field. - Provide a local screenshot/selector review screen rather than proxying arbitrary authenticated pages into an iframe. Require authorization and bounded refresh intervals; no captcha bypass. - Keep secrets outside client bundles and exported projects; document what leaves the device and make retention/deletion controls visible. ## Agent implementation rules - Project rule — domain: Store RecipeVersion, PageSnapshot, RowIdentity and Observation; a stable product key separates a price change from a reordered table. Empty extraction is an error until reviewed. - Project rule — scope and recovery: Provide a local screenshot/selector review screen rather than proxying arbitrary authenticated pages into an iframe. Require authorization and bounded refresh intervals; no captcha bypass. - Project rule — acceptance: Reorder rows, change a price and remove the price selector; report one price change for the right product, followed by a failed observation rather than all products disappearing. - Project rule — delivery: document real setup commands and permissions; do not claim a build, accuracy level, performance result or security certification that has not been demonstrated. ## Optional agent skills and references - Recommended skill: [modern-python](https://github.com/trailofbits/skills/blob/main/plugins/modern-python/skills/modern-python/SKILL.md) — structure the Python worker or explicitly optional read-only utility with pinned dependencies, typed boundaries and clear failure handling. Follow the maintainer's installation instructions and match its requirements to the chosen runtime. - Recommended skill: [agent-browser](https://github.com/vercel-labs/agent-browser/blob/main/skills/agent-browser/SKILL.md) — inspect the approved browser workflow and reproduce permitted page interactions; the CLI/browser runtime is a separate prerequisite. Follow the maintainer's installation instructions and match its requirements to the chosen runtime. Read the linked SKILL.md and its dependencies before adding a skill. Select only the skills matching this project's runtime and task; their documentation does not supply API access, credentials or approval to perform external actions. Pin the reviewed revision where the tool supports it. Follow the chosen agent's documented project-level installation mechanism. ## Distribution ideas These are optional planning notes. Obtain the owner's approval before publishing or contacting anyone. - Demonstrate this working slice using synthetic or explicitly authorized non-sensitive examples: Save a reviewed selector recipe for one catalogue table, preview extracted fields, schedule refreshes and show field-level changes. - Share a synthetic example export and the acceptance walkthrough; keep real customer, health, financial and source data private: Reorder rows, change a price and remove the price selector; report one price change for the right product, followed by a failed observation rather than all products disappearing. - State the limits before asking someone to replace their existing tool: Provide a local screenshot/selector review screen rather than proxying arbitrary authenticated pages into an iframe. Require authorization and bounded refresh intervals; no captcha bypass. ## Engineering roadmap 1. Phase 1 — Pin the working slice and create its example input: Save a reviewed selector recipe for one catalogue table, preview extracted fields, schedule refreshes and show field-level changes. Confirm setup: A small authorized URL set, reviewed selectors or field schema and a user-approved crawl frequency; no proxy-evasion machinery. 2. Phase 2 — Implement persistence and write-time invariants before decorating the UI: Store RecipeVersion, PageSnapshot, RowIdentity and Observation; a stable product key separates a price change from a reordered table. Empty extraction is an error until reviewed. 3. Phase 3 — Connect the working view to real saved state. Bound hosts, redirects, page count, body size and timeout. Reject private-network destinations throughout redirects, respect access restrictions and backoff, and distinguish extraction failure from a missing field. 4. Phase 4 — Expose the app-specific limits and recovery path in context: Provide a local screenshot/selector review screen rather than proxying arbitrary authenticated pages into an iframe. Require authorization and bounded refresh intervals; no captcha bypass. 5. Phase 5 — Walk through this concrete acceptance case and preserve its exported evidence: Reorder rows, change a price and remove the price selector; report one price change for the right product, followed by a failed observation rather than all products disappearing. Finish the README and backup/restore instructions; report unfinished capabilities explicitly. ## Paid-product capabilities outside this build - OAuth app verification - durable execution at scale - schema drift handling and enterprise controls - hundreds of maintained connectors ## Implementation prompt Build the following focused alternative to Browse AI. This is a deliberately limited personal or small-team substitute, not parity with the paid service. WORKING SLICE Save a reviewed selector recipe for one catalogue table, preview extracted fields, schedule refreshes and show field-level changes. SETUP AND ARCHITECTURE Use Python, HTTP parsing, SQLite and optional Playwright for explicitly permitted rendered pages. Prerequisites: A small authorized URL set, reviewed selectors or field schema and a user-approved crawl frequency; no proxy-evasion machinery. Before integrating anything, record actual versions and permissions, plus model files or provider limits only where used, in the README; make unavailable dependencies visible rather than simulating success. DOMAIN MODEL AND INVARIANTS Store RecipeVersion, PageSnapshot, RowIdentity and Observation; a stable product key separates a price change from a reordered table. Empty extraction is an error until reviewed. IMPLEMENTATION CONTRACT Bound hosts, redirects, page count, body size and timeout. Reject private-network destinations throughout redirects, respect access restrictions and backoff, and distinguish extraction failure from a missing field. Provide an input/setup view, the main work view, and a review/export view appropriate to this workflow. Preserve the last saved state if a job or save fails. Include empty, loading, permission-denied, partial and retryable-error states. Log identifiers and error categories without secret values or unnecessary private content. APP-SPECIFIC BOUNDARY AND RECOVERY Provide a local screenshot/selector review screen rather than proxying arbitrary authenticated pages into an iframe. Require authorization and bounded refresh intervals; no captcha bypass. ACCEPTANCE SCENARIO Reorder rows, change a price and remove the price selector; report one price change for the right product, followed by a failed observation rather than all products disappearing. Also reopen the app after an interrupted operation, confirm the saved record/export remains inspectable, and document the recovery action. These are implementation acceptance requirements, not a claim that this guide has been tested. DELIVERY Deliver a runnable repository with migrations or project-format versioning, a non-sensitive example, environment/permission setup, the exact manual acceptance steps, and a backup/export-and-restore walkthrough. Implement the working slice before optional integrations; list any deferred paid-product capabilities honestly. Do not add capabilities outside the working slice just to resemble the original product. PROJECT RULES FOR AGENTS.md Keep the domain invariants above executable at the write boundary. Propose scope changes before adding providers or permissions. Never fabricate source evidence, publish results, identity matches or successful delivery. Preserve user originals and require an explicit confirmation for destructive changes or external publication. ## Completion evidence Demonstrate the prompt's acceptance scenarios against the scoped workflow. Include setup from a clean checkout and failure recovery. Check persistence across restart and export/restore only for the state the prompt says to store; for memory-only tools, confirm that temporary content is discarded as specified. Record actual results and remaining limitations. A detailed plan alone does not establish a working replacement.