Skip to content

presearch.md

Latest commit

 

History

History
83 lines (62 loc) · 3.55 KB

presearch.md

File metadata and controls

83 lines (62 loc) · 3.55 KB
Raw

connector-presearch

Presearch decentralized search engine integration. Search and scrape actions with TTL-based result caching.

       Rule engine                    connector-presearch              Presearch
            │                                 │                            │
            │ execute(search, {query})        │                            │
            ├────────────────────────────────►│                            │
            │                                 │ cache lookup               │
            │                                 │ (TTL = cache_ttl_secs)     │
            │                                 │                            │
            │                                 │ miss ──────────────────────┤
            │                                 │  Bearer <api_key>          │
            │                                 │ HTTPS                      │
            │                                 │  ◄─────────────────────────┤
            │                                 │                            │
            │                                 │ store in cache             │
            │  results                        │                            │
            │  ◄──────────────────────────────┤                            │
            │                                 │                            │
            │ execute(scrape, {url})          │                            │
            ├────────────────────────────────►│ check allowed_scrape_hosts │
            │                                 │  ◄─────────────────────────┤

Fig. 1. Presearch data flow. Search results are cached with a configurable TTL; scrape targets must match allowed_scrape_hosts before a request is made.

1. Configuration

TABLE I. CONFIG FIELDS

Field Type Default Description
api_key Secret<String> (required) Presearch API key
api_base String "https://presearch.com" Presearch API base URL
cache_ttl_secs u64 300 (5 minutes) Cache TTL for search/scrape results
allowed_scrape_hosts Vec<String> [] Additional hosts allowed for scrape action

2. Authentication

API key passed as a header on each request.

3. Triggers

None. This is an action-only connector.

4. Actions

TABLE II. ACTIONS

Name Input fields Output fields
search query: String results: Object, cached: bool
scrape url: String content: String, cached: bool

Both actions use the same TTL-based cache. Repeated identical requests within the TTL window return cached results (indicated by cached: true).

5. Capabilities Required

Capability Parameter
NetworkOutbound presearch.com
NetworkOutbound Each host in allowed_scrape_hosts

6. Example Rule

[rule]
name = "scheduled-search"

[trigger]
type = "Cron"
expression = "0 9 * * *"

[[actions]]
type = "RunConnector"
connector = "connector-presearch"
action = "search"

[actions.params]
query = "Springtale automation platform"