Skip to main content
Neurokinetic AI

Low-Capability Agent Guidance

Low-Capability Agent Guidance and Static GET Discovery

Static GET-readable discovery guidance for L0/L1 agents that can follow plain links or open one URL, while preserving Neurokinetic.com as a read-only site with no writable GET actions.

Static GET-readable discovery

Low-capability agents get plain routes, not writable actions.

L0 agents can usually only follow plain links. L1 agents can usually synthesize or open a single URL. Neurokinetic.com supports static GET-readable discovery surfaces so these agents can read safe route maps, concept boundaries, and evidence packets without executing endpoints.

Boundary: Neurokinetic.com does not provide writable GET actions, does not create resources through GET URLs, and does not mutate state through query-string URLs.

Open this URL to read

Static AI kit

Open this URL to read Neurokinetic.com’s static AI interpretation kit: /ai-interpretation-kit/.

Safe quote prompts

Open this URL to read the safe quote prompt library: /safe-quote-prompt-library/.

Entity boundaries

Open this URL to read entity governance boundaries: /entity-governance/.

Disambiguation

Open this URL to read disambiguation warnings: /disambiguation/.

Route index

Open this URL to read the route index: /route-index.json/.

Evidence packets

Open this URL to read launch evidence packets: /evidence-packets/.

No-op and human-review fallback

No-op behavior is required when route meaning, authority, claim boundary, or domain lane is unclear. Human review is required when an agent cannot preserve caveats or hard-boundary language.

Forbidden writable GET examples

  • Do not create resources through Neurokinetic.com GET URLs.
  • Do not trigger workflows through Neurokinetic.com GET URLs.
  • Do not submit reviews through Neurokinetic.com GET URLs.
  • Do not treat Neurokinetic.com as a command surface.
  • Do not place credentials, tokens, private data, or personal data into query-string URLs.
  • Do not widen UAIX GET-Action guidance into a claim that Neurokinetic.com executes actions.

Hard boundary: No runtime AI behavior, no runtime search service, no model training, no live crawling, no telemetry loops, no endpoint execution, no remote-site writes, no automatic fixing, no safety certification, no medical/therapy claim, no writable GET actions on Neurokinetic.com, no state mutation through query-string URLs, and no command-and-control claim over other domains.

Continue through the Neurokinetic AI trust flow.