Skip to content

cli-reference.md

Latest commit

 

History

History
86 lines (78 loc) · 5.28 KB

cli-reference.md

File metadata and controls

86 lines (78 loc) · 5.28 KB

goresearch CLI reference

This page is auto-generated from the CLI flag definitions.

Usage

goresearch [flags]
goresearch init
goresearch doc

Flags

  • -cache.clear (default: false) — Clear cache directory before run
  • -cache.dir (default: .goresearch-cache) — Cache directory path
  • -cache.maxAge (default: 0s) — Max age for cache entries before purge (e.g. 24h, 7d); 0 disables
  • -cache.strictPerms (default: false) — Restrict cache permissions (0700 dirs, 0600 files)
  • -cache.topicHash (default: ``) — Optional topic hash to scope cache; accepted for traceability
  • -debug-verbose (default: false) — Allow logging raw chain-of-thought (CoT) for debugging Harmony/tool-call interplay
  • -domains.allow (default: ``) — Comma-separated allowlist of hosts/domains; if set, only these are permitted (subdomains included)
  • -domains.deny (default: ``) — Comma-separated denylist of hosts/domains; takes precedence over allow
  • -dry-run (default: false) — Plan and select without calling the model
  • -enable.pdf (default: false) — Enable optional PDF ingestion (application/pdf)
  • -input (default: request.md) — Path to input Markdown research request
  • -lang (default: ``) — Optional language hint, e.g. 'en' or 'fi'
  • -llm.base (default: ``) — OpenAI-compatible base URL
  • -llm.key (default: ``) — API key for OpenAI-compatible server
  • -llm.model (default: ``) — Model name
  • -max.perDomain (default: 3) — Maximum sources per domain
  • -max.perSourceChars (default: 12000) — Maximum characters per source extract
  • -max.sources (default: 12) — Maximum number of sources
  • -min.snippetChars (default: 0) — Minimum non-whitespace snippet characters to keep a result (0 disables)
  • -output (default: report.md) — Path to write the final Markdown report
  • -robots.overrideConfirm (default: false) — Second confirmation flag required to activate robots override allowlist
  • -robots.overrideDomains (default: ``) — Comma-separated domain allowlist to ignore robots.txt (use with --robots.overrideConfirm)
  • -search.file (default: ``) — Path to JSON file for offline file-based search provider
  • -searx.key (default: ``) — SearxNG API key (optional)
  • -searx.ua (default: goresearch/1.0 (+https://github.com/hyperifyio/goresearch)) — Custom User-Agent for SearxNG requests
  • -searx.url (default: ``) — SearxNG base URL
  • -synth.systemPrompt (default: ``) — Override synthesis system prompt (inline string)
  • -synth.systemPromptFile (default: ``) — Path to file containing synthesis system prompt
  • -tools.dryRun (default: false) — Do not execute tools; emit dry-run envelopes
  • -tools.enable (default: false) — Enable tool-orchestrated chat mode
  • -tools.maxCalls (default: 32) — Max tool calls per run
  • -tools.maxWallClock (default: 0s) — Max wall-clock duration for tool loop (e.g. 30s); 0 disables
  • -tools.mode (default: harmony) — Chat protocol mode: harmony|legacy
  • -tools.perToolTimeout (default: 10s) — Per-tool execution timeout (e.g. 10s)
  • -v (default: false) — Verbose logging
  • -log.level (default: ``) — Structured log level for file output: trace|debug|info|warn|error|fatal|panic (default info)
  • -log.file (default: ``) — Path to write structured JSON logs (default logs/goresearch.log)
  • -verify.systemPrompt (default: ``) — Override verification system prompt (inline string)
  • -verify.systemPromptFile (default: ``) — Path to file containing verification system prompt
  • -verify/-no-verify (default: -verify) — Enable or disable the fact-check verification pass and Evidence check appendix

Environment variables

  • LLM_BASE_URL: OpenAI-compatible base URL
  • LLM_MODEL: Model name
  • LLM_API_KEY: API key
  • SEARX_URL: SearxNG base URL (or SEARXNG_URL)
  • SEARX_KEY: SearxNG API key (or SEARXNG_KEY)
  • CACHE_DIR: Cache directory path
  • LANGUAGE: Language hint
  • SOURCE_CAPS: Max sources and optional per-domain cap as '' or ','
  • CACHE_MAX_AGE: Purge cache entries older than this duration (e.g. 24h, 7d)
  • DRY_RUN: Enable dry-run when truthy
  • VERBOSE: Enable verbose logs when truthy
  • CACHE_CLEAR: Clear cache before run when truthy
  • CACHE_STRICT_PERMS: Restrict cache permissions when truthy
  • HTTP_CACHE_ONLY: Serve HTTP bodies only from cache; fail on miss
  • LLM_CACHE_ONLY: Serve LLM results only from cache; fail on miss
  • ROBOTS_OVERRIDE_DOMAINS: Comma-separated allowlist to ignore robots.txt; requires robots.overrideConfirm
  • DOMAINS_ALLOW: Comma-separated allowlist of hosts/domains
  • DOMAINS_DENY: Comma-separated denylist of hosts/domains
  • SYNTH_SYSTEM_PROMPT: Inline synthesis system prompt override
  • SYNTH_SYSTEM_PROMPT_FILE: Path to synthesis system prompt file
  • VERIFY_SYSTEM_PROMPT: Inline verification system prompt override
  • VERIFY_SYSTEM_PROMPT_FILE: Path to verification system prompt file
  • VERIFY: Set to truthy to force enable verification (overrides NO_VERIFY)
  • NO_VERIFY: Set to truthy to disable verification
  • LOG_LEVEL: Structured log level for file output (trace|debug|info|warn|error|fatal|panic)
  • LOG_FILE: Path to write structured JSON logs (default goresearch.log)
  • TOPIC_HASH: Optional topic hash to scope cache

Generated by goresearch doc.