{"data":{"kind":"file","path":"README.md","version_id":"bdsw4nci69kibcsvqel43nmg","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":3083,"modified_at":"2026-11-09T09:38:43.643000","content_hash":"44e7de59cfc5d7bcda124cad1f8d9d5efd4468008417fe402f36923a0f23783a"},"entries":[],"content":"# navital-gst\n\n**GST Desk**, one of six worlds in [Navital Gym](https://github.com/Navitallabs/navital-gym) v1.1.0, as a verifiers environment. By [Navital Labs](https://navitallabs.com).\n\n- **What:** 50 public tasks. An agent does GST invoicing and return computations from documents under written, simplified rules: invoices (CGST/SGST/IGST, round-off), place of supply, ITC reconciliation, late filing (interest and fee), and registration thresholds. The world has 5 tools (calendar, documents, policy, rate lookup, GSTIN validation).\n- **Grading:** deterministic rules, never an AI judge. Reward = the world's own grader (score = fraction of rules passed, in [0, 1]) run on the last JSON object in the final message. `solved` (all required rules passed) is a second metric. No GST task has a forbidden step, so a forbidden-step column does not apply here.\n- **Hidden fields** (expected answers, traps) never reach the agent or the traces, as long as the agent only has the world's tools; they are read host-side at scoring time. A runtime that gives the agent a shell or file access (for example a hosted sandbox with a shell) has not been tested.\n- **Public only:** the same public tasks as the open Gym. Models may have seen them. Our private held-out sets are NOT in this package.\n- **Synthetic data and simplified rules.** Not tax advice and not a statement about current GST law or GSTN systems. Not affiliated with GSTN or any tax authority.\n\n## Use\n```\nprime env install navitallabs/navital-gst          # once published\nvf-eval navital-gst -m <model> -n 50\n```\n```python\nfrom navital_gst import load_environment\nenv = load_environment()                 # all 50 tasks; ids=[\"IN-GST-001\"] for a subset\n```\n**Runtime (read this):** the per-rollout tool server needs `navital-gym` on the machine that runs it, so use a local runtime, for example:\n```\nvf-eval navital-gst -m <model> --env.agent.runtime.type subprocess --env.agent.harness.id null\n```\nPrime's default hosted sandbox runtime (a bare python:3.11-slim image) has NOT been tested with this package and is not expected to work without navital-gym in the image.\n\nNeeds `navital-gym==1.1.0` importable (a dependency). The per-rollout tool server runs as a local subprocess (`python -m navital_gym.adapters.verifiers_env`).\n\n## Honest limits\nPrompt: tasks carry the Gym's own prompt including the \"Answer fields\" line (0.1.2 fix: version 0.1.1 and the generic navital-gym 1.1.0 adapter drop that line for GST, so a model does not know the `tax_type` and other fields the grader needs).\nWritten against verifiers 0.4.0's v1 API (pinned), which is young. Tested offline with a scripted model and by a reward-parity test (reward and `solved` equal the Gym's own grader on 250 recorded real rollouts, 5 models x 50 tasks); not run on Prime's hosted evals.\n\n## Results\nReference results are on the public leaderboard: https://navitallabs.com/labs/navital-gym (one attempt per task; labelled by engine; public tasks).\n\n## Licence\nCode: Apache-2.0 (LICENSE). Task data: CC BY 4.0, attribution to Navital Labs / Navital Gym.\n","encoding":"utf-8","truncated":false,"total_bytes":3083},"status":null}