Challenge · 01
Mode honesty
live, mock, and degraded runs must stay visibly distinct.
<- AI Systems Dossier
ENTRY NO. 04Sales Technology / R&DPrototypeTinyFish / TypeScript / SQLite replayAGENT WORKFLOW / PROTOTYPE
Scoutbound
Identity · Agentic prospecting capability demo
Agentic prospecting workflow. Research leads, keep evidence, score fit, export to CRM.
TinyFish / TypeScript / SQLite replay
Evidence boundary
What is demonstrated, and where the claim stops.
Repository factConfiguration
What this proves
- /The workflow makes live, mock, and degraded execution states explicit.
- /Session replay can preserve evidence, scoring output, and CRM handoff state.
- /Versioned JSON/CSV exports provide a reviewable boundary.
What this does not prove
- /Benchmark-quality extraction accuracy.
- /A measured speed improvement over manual research.
- /A persistent monitored service or production CRM deployment.
Data boundary
Mock mode is the safe default. Live TinyFish calls require explicit configuration; replay fixtures and mock output are labelled.
Evidence type
repository fact / configuration
Mock mode is the safe default. Live browser-agent performance and extraction quality are not yet benchmarked publicly.
Known limitations
- /Mock mode is the default.
- /Live execution depends on a paid external browser-agent service.
- /Two historical repositories need consolidation around scoutbound_agent.
What production would require
- /A persistent managed database and deployed API.
- /A labelled extraction evaluation set and saved benchmark results.
- /Tenant controls, secret management, monitoring, and CRM integration testing.
Problem
Browser-agent output needs explicit execution modes, persisted evidence, and a safe handoff boundary.
System Flow
5 stages · click to inspect
->
->
->
->
Click any stage to inspect
Signature Module · Agentic prospecting capability demo
Replay the workflow boundary
See how execution mode, evidence, uncertainty, score reasons, persistence, and export remain visible.
Scoutbound · explicit execution modeinspectable
MODElive · mock · degraded
BRIEFmarket · company size · keywords · role
AGENTTinyFish discovery and inspection adapters
TRACEevidence · uncertainty · score reasons
OUTPUTpersisted session · JSON / CSV handoff
Demonstrated Change
Demo scopeBefore
Disconnected prospect research and copy-paste handoff.
->
Demonstrated build
A replayable prototype with explicit modes, evidence, scoring, and versioned export.
The prototype demonstrates a reviewable path from prospecting brief to evidence-bearing lead export.
Mock mode is the safe default. Live browser-agent performance and extraction quality are not yet benchmarked publicly.
Proof Signals
Classified evidence3
run modes
Repository fact
2
export formats
Repository fact
1
CRM handoff contract
Repository fact
Every signal is classified as a repository fact, recorded test, configuration, or disclosed data boundary.
Why This Is Hard
3 engineering challengesChallenge · 02
Evidence retention
output must preserve sources and uncertainty.
Challenge · 03
Safe external calls
paid traffic and CRM writes require explicit opt-in.
Engineering Depth
3 topics · click to expandThe canonical repository blocks live calls unless explicitly enabled and keeps mock or degraded output labelled.
Key numbers
Built withTinyFishTypeScriptExpressReactZodSQLite