feat: hermes models / hermes providers — non-interactive inventory CLI (#23359) by kshitijk4poor · Pull Request #23369 · NousResearch/hermes-agent

kshitijk4poor · 2026-05-10T19:08:50Z

What this PR does

Adds three scriptable, non-interactive CLI surfaces and consolidates four existing inventory call-sites onto a single source of truth:

hermes models list   [--provider NAME] [--json] [--all] [--no-live] [--offline]
hermes models status [--provider NAME] [--json]                     [--offline]
hermes providers list                  [--json] [--all]             [--offline]

Closes #23359 (jointly with #3503 once that PR migrates to consume the shared formatter).

The new hermes_cli/inventory.py module (600 LOC) is the single source of truth for inventory traversal. The dashboard (/api/model/options), TUI (model.options JSON-RPC), and TUI (model.save_key JSON-RPC) have been migrated to consume it, dropping ~140 lines of duplicated config-load / canonical-merge / picker-hint logic across the three consumer files. --offline mode synthesises from in-repo static data so cron and CI callers don't block on the network.

A reachability-probe surface (--probe for models status) is tracked separately in #23614 — provider_model_ids() falls back to curated lists, so a true reachability check needs an explicit endpoint probe that isn't wired through the existing model-discovery wrapper.

Why

Hermes had four independent ways to list providers/models — each interactive (TTY required) or HTTP (gateway running), each reproducing the same five-step config dance, and each post-processing the result a slightly different way. There was no scriptable / cron-friendly entry point.

Surface	Before	After
Interactive picker (`cli.py`)	`list_authenticated_providers` directly	(unchanged)
Dashboard (`/api/model/options`)	Inline 17-line config dance + raw `cfg.get("custom_providers")` (which misses v12+ keyed providers — bug fix) + `list_authenticated_providers`	`inventory.build_payload("models", ctx=load_picker_context(), max_models=50)`
TUI (`model.options` JSON-RPC)	Inline 17-line config dance + `list_authenticated_providers` + 45-line canonical-merge / picker-hint post-processing	`inventory.build_payload("models", ctx, picker_hints=True, canonical_order=True)` with `with_overrides()` for agent-session state
TUI (`model.save_key` JSON-RPC)	Inline `list_authenticated_providers` + slug match	`inventory.build_payload("models", ctx, picker_hints=True)` then slug match
NEW: scriptable CLI	(didn't exist)	`hermes models list/status` + `hermes providers list`

Five open PRs have each reinvented the same traversal for one narrow purpose: #3503 (/models slash), #21695 (plugin provider runtime resolver), #19526 (model.picker_mode), #17994 (/free-models), #21785 (hermes doctor skip-by-endpoint). Each adds ~50-150 LOC of inventory walking. Now they have a single helper to consume.

Full motivation, pain points, and ecosystem evidence in tracking issue #23359.

What this PR is NOT doing

Not refactoring list_authenticated_providers — only delegating to it. No behaviour change.
Not changing the existing hermes model interactive picker.
Not adding write-side symmetry for hermes model → providers.<slug>.models — that's Make providers the single source of truth for model registry #16622 (separate ~50 LOC follow-up).
Not implementing model.allowlist / model.picker_mode — that's feat(cli): add picker_mode config to restrict model picker #19526.
Not adding model-discovery for built-in providers — that's [Feature]: Dynamic Model Discovery and Filtering for Built-in Providers #13055 / fix(providers): wire API-key model-provider plugins into runtime and /model #21695.
Not adding a reachability-probe — that's follow-up: hermes models status --probe (reachability check, extracted from #23369) #23614.

Modes / network behaviour

Flag	`models.dev` HTTP	Per-provider live HTTP
(default)	✓ via cache	✓ for configured providers
`--no-live` (`models list` only)	✓ via cache	✗
`--offline`	✗	✗

--offline bypasses list_authenticated_providers entirely and synthesises inventory from CANONICAL_PROVIDERS (universe of ~35) + _PROVIDER_MODELS (curated) + OPENROUTER_MODELS (local constant) + profile.fallback_models. Auth state is derived from environment variables alone — profile.env_vars ∪ PROVIDER_REGISTRY[slug].api_key_env_vars ∪ HERMES_OVERLAYS[slug].extra_env_vars. No consultation of auth.json, credential pool, Claude Code creds, or Hermes OAuth files.

Public API exposed by `hermes_cli/inventory.py`

load_picker_context() -> ConfigContext
    # Replaces the 17-LOC config dance in dashboard / TUI / picker.

ConfigContext.with_overrides(*, current_provider, current_model, current_base_url) -> ConfigContext
    # Layer agent-session state on top of disk config (TUI override path).

build_payload(kind, *, ctx=None, provider=None, include_unconfigured=False,
              live=True, offline=False, picker_hints=False,
              canonical_order=False, max_models=50) -> dict
    # kind ∈ {"models", "status", "providers"}

Optional flags:

ctx — pre-loaded context, bypassing load_picker_context() (TUI override path).
picker_hints=True — adds authenticated per row, plus auth_type / key_env / warning for unconfigured canonical rows (the dashboard frontend's ModelPickerDialog shape).
canonical_order=True — reorder rows to CANONICAL_PROVIDERS declaration order (TUI display order).

authenticated is only emitted under picker_hints=True. Default scriptable shape uses auth_state: "configured" | "unconfigured" exclusively — clean schema for cron/CI scripts.

JSON contract (default scriptable shape)

{
  "schema_version": 1,
  "current": {"provider": "openrouter", "model": "anthropic/claude-sonnet-4.6"},
  "providers": [
    {
      "slug": "anthropic",
      "name": "Anthropic",
      "is_current": false,
      "is_user_defined": false,
      "auth_state": "configured",
      "api_mode": "anthropic_messages",
      "auth_type": "api_key",
      "base_url": "https://api.anthropic.com",
      "env_vars": ["ANTHROPIC_API_KEY", "ANTHROPIC_TOKEN", "CLAUDE_CODE_OAUTH_TOKEN"],
      "env_vars_present": [true, false, false],
      "models": ["claude-opus-4-7", "..."],
      "total_models": 8,
      "source": "canonical",
      "model_source": "curated"
    }
  ]
}

env_vars lists variable names (including HERMES_OVERLAYS bridges — e.g. OpenRouter accepts OPENAI_API_KEY). env_vars_present is the corresponding bool array. Values are never included in the payload — explicit invariant; canary-tested.

Substrate gotchas the implementation respects

These are documented in references/provider-inventory-surface.md (hermes-agent-dev skill) and were each verified against main before implementation:

Four registries don't fully overlap. Inventory iterates CANONICAL_PROVIDERS (the universe, ~35).
profile.fetch_models() returns models the runtime rejects. OpenRouter raw=367 vs picker=34. Module never calls it directly. Canary test enforces.
Custom providers split across two config shapes. Module reads via get_compatible_custom_providers(config) — fixes a latent bug in dashboard's /api/model/options (was missing v12+ keyed entries).
OpenRouter offline gotcha. _PROVIDER_MODELS["openrouter"] is None; module uses OPENROUTER_MODELS constant. ~31 entries, no HTTP. Canary asserts equality against the constant length.
CLI exit-code propagation. hermes_cli/main.py:11688 does args.func(args) and discards the return value. Handlers use parser.error() / required=True / SystemExit(2) instead of return 2.
TUI contract preserved: model.options does NOT call provider_model_ids per-row in the live path. That would pull non-agentic models (TTS, embeddings). Canary test (test_live_path_does_not_call_models_for) enforces.
model.save_key migration verified. build_payload("models", picker_hints=True) returns the freshly-keyed slug with populated models. Canary test (test_save_key_migration_returns_populated_row) enforces.
Anthropic credential discovery has 5+ paths. A clean-env E2E asserting "anthropic is unconfigured" can fail on dev machines with Claude Code installed. E2E uses arcee (single env var, no extras) as the canonical clean-baseline target.

Tests

File	Tests	What
`tests/hermes_cli/test_inventory.py`	32	Three concept canaries (no models.dev / no provider HTTP / status HTTP-free) + H2-H6 + W1 regressions + alias canonicalisation + universe assertions + schema/mutex + consumer-shape parity (web_server + TUI) + TUI live-path canary + save_key migration canary
`tests/hermes_cli/test_inventory_cli.py`	9	Subprocess E2E via `hermes` console script
`tests/test_tui_gateway_server.py::test_model_options_does_not_overwrite_curated_models`	1	Existing test, still passes (TUI contract gate)

Concept B canary monkeypatches every per-provider live-fetch helper (fetch_anthropic_models, fetch_openrouter_models, fetch_ai_gateway_models, fetch_ollama_cloud_models, fetch_lmstudio_models, _fetch_github_models, fetch_api_models, fetch_nous_models, fetch_models_dev, model_ids, get_curated_nous_model_ids, bedrock_model_ids_or_none, list_authenticated_providers) to fail-on-call, then asserts none fire when --offline is passed.

The TUI contract canary monkeypatches _models_for and asserts it fires zero times in the live path with picker_hints=True, canonical_order=True (the TUI's exact call shape) — guaranteeing build_payload doesn't pull non-agentic models when consumed via model.options.

The save_key migration canary monkeypatches list_authenticated_providers to return a fake row, then asserts the saved slug is found in build_payload's output with populated models — locking the contract tui_gateway/server.py:5247-5256 relies on.

$ pytest tests/hermes_cli/test_inventory.py tests/hermes_cli/test_inventory_cli.py \
        tests/test_tui_gateway_server.py::test_model_options_does_not_overwrite_curated_models -q
42 passed in 5.8s

Full tests/hermes_cli/ + tests/test_tui_gateway_server.py: 13 pre-existing failures verified to fail identically on origin/main (gateway systemd / WSL / kanban / plugins / web-server auth / update hangup / browser launch hint — none related to this PR).

Lint diff

ruff — Total: 0 on HEAD, 0 on base (➖ 0)  🆕 New issues: none
ty   — 0 issues in hermes_cli/inventory.py

Files changed

hermes_cli/inventory.py                | 600 +++++++++++++++++++ (new, single source of truth)
hermes_cli/main.py                     |  89 ++ (subparsers + 3 handlers + #23614 TODO marker)
hermes_cli/web_server.py               |  36 +- (-29 net: migrated to inventory)
tui_gateway/server.py                  | 155 +- (migrated model.options + model.save_key)
tests/hermes_cli/test_inventory.py     | 678 ++++ (new)
tests/hermes_cli/test_inventory_cli.py | 136 ++ (new)
website/docs/reference/cli-commands.md |  84 +++
7 files changed, 1638 insertions(+), 140 deletions(-)

Of the 140 lines deleted from web_server.py + tui_gateway/server.py, ~125 were literal duplicates of the config dance + canonical merge, now living in one place.

No setup.py files. No env-var-value logging (canary-tested). No changes to list_authenticated_providers, provider_model_ids, or any provider plugin.

How to verify

# Universe + offline correctness
hermes providers list --offline --all --json | jq '.providers | length'           # 35
hermes models list --provider=openrouter --offline --json | jq '.providers[0].total_models'   # ≈31

# Default mode (live, same as picker)
hermes providers list --json | jq '.providers | length'                            # configured providers count

# OpenRouter regression check
hermes models list --provider=openrouter --json | jq '.providers[0].total_models'  # ≤50, NOT 367

# Mutex + exit codes
hermes models list --provider=does-not-exist; echo $?   # 2
hermes models foo; echo $?    # 2
hermes providers; echo $?     # 2

# Dashboard / TUI integration: their existing endpoints work unchanged
curl http://localhost:7172/api/model/options | jq '.providers | length'   # ≥1
# (TUI model.options + model.save_key exercised by tests/test_tui_gateway_server.py)

Out of scope (tracked separately)

follow-up: hermes models status --probe (reachability check, extracted from #23369) #23614 — --probe reachability flag for models status. TODO marker added at the argparse site.
Make providers the single source of truth for model registry #16622 — write-side symmetry for hermes model → providers.<slug>.models (~50 LOC follow-up).
Substrate audit — empirical findings in tracking: provider/model inventory has no scriptable surface — five PRs reinvent it, four issues blocked #23359 (registry overlaps, profile.fetch_models() unfiltered, Nous offline gap, list_authenticated_providers unconditional fetches) — separate audit issue once this lands.
feat(gateway): add /models slash command for model discovery #3503 — /models slash unification. After merge, comment proposing migration to inventory.build_payload.
CLI picker (hermes model) consumption. Interactive-only, has filtering logic (drops empty rows) that doesn't belong on a scriptable surface.

Refs

Tracking issue: tracking: provider/model inventory has no scriptable surface — five PRs reinvent it, four issues blocked #23359
Probe follow-up: follow-up: hermes models status --probe (reachability check, extracted from #23369) #23614
Substrate: feat(providers): ProviderProfile ABC + plugins/model-providers/ (salvage of #14424 + pluggable migration) #20324 (merged 2026-05-05)
Original refactor: feat: provider modules — ProviderProfile ABC, 29 providers, fetch_models, transport single-path #14424
Issues this helps close: [Feature]: /models slash command for listing available models #3500, [Feature]: Dynamic Model Discovery and Filtering for Built-in Providers #13055
Open PRs that should consume the new surface: feat(gateway): add /models slash command for model discovery #3503, feat(cli): add picker_mode config to restrict model picker #19526, fix(providers): wire API-key model-provider plugins into runtime and /model #21695, feat(cli): add /free-models command to list free OpenRouter models #17994, fix(doctor): skip /models health check for providers that don't support it #21785

Consolidates the duplicated provider/model traversal across dashboard /api/model/options, TUI model.options, and TUI model.save_key onto a single hermes_cli/inventory.py module (600 LOC). The 3 consumer files delete 150 LOC of inline config-slice + post-processing. Also adds three scriptable CLI surfaces for cron / CI / agents introspecting their own environment: hermes models list [--provider NAME] [--json] [--all] [--no-live] [--offline] hermes models status [--provider NAME] [--json] [--offline] hermes providers list [--json] [--all] [--offline] --offline is fully HTTP-free (no models.dev, no provider live API, no remote catalog fetches, no credential-pool probing); auth state derived from environment variables alone via profile.env_vars ∪ PROVIDER_REGISTRY.api_key_env_vars ∪ HERMES_OVERLAYS.extra_env_vars. Reachability-probe surface (--probe) is tracked separately in NousResearch#23614. Closes NousResearch#23359.

kshitijk4poor · 2026-05-11T08:28:30Z

Closing in favour of #23666, which is the cut-to-the-bone refactor extracted from this PR.

This PR (#23369) bundled two concerns:

The consolidation — three call-sites (dashboard /api/model/options, TUI model.options, TUI model.save_key) each duplicating a 17-LOC config-slice + list_authenticated_providers + post-processing. Real win.
A new scriptable CLI surface — hermes models list/status + hermes providers list with offline mode, JSON contract, text renderers. No external user demand — none of the cited "five PRs reinventing this traversal" actually need or would consume the new helper, and no other issue in the tracker asks for hermes models list --json.

Honest accounting on this PR was +1638 / -140 with a 0.23 consolidation ratio (additions in the new module vs deletions in consumers) — well below the 0.5 threshold for legitimate consolidation framing. ~1400 of those LOC served (1) — speculative feature work I retroactively justified as part of the consolidation argument.

#23666 keeps only (1):

240 LOC module (vs 600 here)
378 LOC tests (vs 814 here)
+666 / -145 total (vs +1638 / -140 here)
Consolidation ratio 0.60 — above threshold

The scriptable CLI surface is deferred until an external user actually asks for it.

Both latent bugs this PR found are preserved in #23666:

Dashboard missing v12+ keyed providers (cfg.get("custom_providers") vs get_compatible_custom_providers).
TUI demoting canonical providers configured via the keyed providers: schema (canonical-order keying on slug membership, not is_user_defined).

Three call-sites in the codebase each duplicated the same config-slice + list_authenticated_providers + post-processing pattern: - hermes_cli/web_server.py /api/model/options - tui_gateway/server.py model.options JSON-RPC - tui_gateway/server.py model.save_key JSON-RPC This consolidates them onto hermes_cli/inventory.py: load_picker_context() -> ConfigContext Replaces the 17-LOC config-slice (model.{default,name,provider, base_url}, providers:, custom_providers:) every consumer did inline. ConfigContext.with_overrides(*, current_provider=, current_model=, current_base_url=) -> ConfigContext Truthy-only overlay for TUI agent-session state on top of disk config. Empty getattr(agent, ...) attrs MUST NOT clobber disk. build_models_payload(ctx, *, include_unconfigured, picker_hints, canonical_order, max_models) -> dict Single payload builder. Delegates curation to list_authenticated_providers (does not call provider_model_ids per row \u2014 that pulls non-agentic models). picker_hints + canonical_order produce the TUI ModelPickerDialog shape; defaults match the dashboard's existing /api/model/options contract. Two latent bugs fixed by consolidation: 1. The dashboard read cfg.get('custom_providers') directly, missing the v12+ keyed providers: form. Now both surfaces go through get_compatible_custom_providers(). 2. The TUI's canonical-merge keyed on is_user_defined to decide order. Section 3 of list_authenticated_providers sets is_user_defined=True on rows from the providers: config dict even when the slug is canonical \u2014 that silently demoted them to the picker tail. _reorder_canonical now keys on slug membership instead. Stats: +666 / -145 (net +521). Module 240 LOC; 18 behavior tests. This PR replaces the rejected #23369 (which bundled the consolidation with new scriptable CLI surfaces \u2014 hermes models list/status, hermes providers list \u2014 and a JSON contract that have no external user demand). Just the refactor; the CLI surface is deferred to a separate PR gated on actual demand. Refs #23359.

Three call-sites in the codebase each duplicated the same config-slice + list_authenticated_providers + post-processing pattern: - hermes_cli/web_server.py /api/model/options - tui_gateway/server.py model.options JSON-RPC - tui_gateway/server.py model.save_key JSON-RPC This consolidates them onto hermes_cli/inventory.py: load_picker_context() -> ConfigContext Replaces the 17-LOC config-slice (model.{default,name,provider, base_url}, providers:, custom_providers:) every consumer did inline. ConfigContext.with_overrides(*, current_provider=, current_model=, current_base_url=) -> ConfigContext Truthy-only overlay for TUI agent-session state on top of disk config. Empty getattr(agent, ...) attrs MUST NOT clobber disk. build_models_payload(ctx, *, include_unconfigured, picker_hints, canonical_order, max_models) -> dict Single payload builder. Delegates curation to list_authenticated_providers (does not call provider_model_ids per row \u2014 that pulls non-agentic models). picker_hints + canonical_order produce the TUI ModelPickerDialog shape; defaults match the dashboard's existing /api/model/options contract. Two latent bugs fixed by consolidation: 1. The dashboard read cfg.get('custom_providers') directly, missing the v12+ keyed providers: form. Now both surfaces go through get_compatible_custom_providers(). 2. The TUI's canonical-merge keyed on is_user_defined to decide order. Section 3 of list_authenticated_providers sets is_user_defined=True on rows from the providers: config dict even when the slug is canonical \u2014 that silently demoted them to the picker tail. _reorder_canonical now keys on slug membership instead. Stats: +666 / -145 (net +521). Module 240 LOC; 18 behavior tests. This PR replaces the rejected NousResearch#23369 (which bundled the consolidation with new scriptable CLI surfaces \u2014 hermes models list/status, hermes providers list \u2014 and a JSON contract that have no external user demand). Just the refactor; the CLI surface is deferred to a separate PR gated on actual demand. Refs NousResearch#23359.

Three call-sites in the codebase each duplicated the same config-slice + list_authenticated_providers + post-processing pattern: - hermes_cli/web_server.py /api/model/options - tui_gateway/server.py model.options JSON-RPC - tui_gateway/server.py model.save_key JSON-RPC This consolidates them onto hermes_cli/inventory.py: load_picker_context() -> ConfigContext Replaces the 17-LOC config-slice (model.{default,name,provider, base_url}, providers:, custom_providers:) every consumer did inline. ConfigContext.with_overrides(*, current_provider=, current_model=, current_base_url=) -> ConfigContext Truthy-only overlay for TUI agent-session state on top of disk config. Empty getattr(agent, ...) attrs MUST NOT clobber disk. build_models_payload(ctx, *, include_unconfigured, picker_hints, canonical_order, max_models) -> dict Single payload builder. Delegates curation to list_authenticated_providers (does not call provider_model_ids per row \u2014 that pulls non-agentic models). picker_hints + canonical_order produce the TUI ModelPickerDialog shape; defaults match the dashboard's existing /api/model/options contract. Two latent bugs fixed by consolidation: 1. The dashboard read cfg.get('custom_providers') directly, missing the v12+ keyed providers: form. Now both surfaces go through get_compatible_custom_providers(). 2. The TUI's canonical-merge keyed on is_user_defined to decide order. Section 3 of list_authenticated_providers sets is_user_defined=True on rows from the providers: config dict even when the slug is canonical \u2014 that silently demoted them to the picker tail. _reorder_canonical now keys on slug membership instead. Stats: +666 / -145 (net +521). Module 240 LOC; 18 behavior tests. This PR replaces the rejected #23369 (which bundled the consolidation with new scriptable CLI surfaces \u2014 hermes models list/status, hermes providers list \u2014 and a JSON contract that have no external user demand). Just the refactor; the CLI surface is deferred to a separate PR gated on actual demand. Refs #23359.

Three call-sites in the codebase each duplicated the same config-slice + list_authenticated_providers + post-processing pattern: - hermes_cli/web_server.py /api/model/options - tui_gateway/server.py model.options JSON-RPC - tui_gateway/server.py model.save_key JSON-RPC This consolidates them onto hermes_cli/inventory.py: load_picker_context() -> ConfigContext Replaces the 17-LOC config-slice (model.{default,name,provider, base_url}, providers:, custom_providers:) every consumer did inline. ConfigContext.with_overrides(*, current_provider=, current_model=, current_base_url=) -> ConfigContext Truthy-only overlay for TUI agent-session state on top of disk config. Empty getattr(agent, ...) attrs MUST NOT clobber disk. build_models_payload(ctx, *, include_unconfigured, picker_hints, canonical_order, max_models) -> dict Single payload builder. Delegates curation to list_authenticated_providers (does not call provider_model_ids per row \u2014 that pulls non-agentic models). picker_hints + canonical_order produce the TUI ModelPickerDialog shape; defaults match the dashboard's existing /api/model/options contract. Two latent bugs fixed by consolidation: 1. The dashboard read cfg.get('custom_providers') directly, missing the v12+ keyed providers: form. Now both surfaces go through get_compatible_custom_providers(). 2. The TUI's canonical-merge keyed on is_user_defined to decide order. Section 3 of list_authenticated_providers sets is_user_defined=True on rows from the providers: config dict even when the slug is canonical \u2014 that silently demoted them to the picker tail. _reorder_canonical now keys on slug membership instead. Stats: +666 / -145 (net +521). Module 240 LOC; 18 behavior tests. This PR replaces the rejected NousResearch#23369 (which bundled the consolidation with new scriptable CLI surfaces \u2014 hermes models list/status, hermes providers list \u2014 and a JSON contract that have no external user demand). Just the refactor; the CLI surface is deferred to a separate PR gated on actual demand. Refs NousResearch#23359.

kshitijk4poor force-pushed the feat/cli-inventory branch from ed732a7 to fd7b94d Compare May 10, 2026 19:26

alt-glitch added type/feature New feature or request comp/cli CLI entry point, hermes_cli/, setup wizard P2 Medium — degraded but workaround exists labels May 10, 2026

kshitijk4poor force-pushed the feat/cli-inventory branch 2 times, most recently from f922f9b to b7f3921 Compare May 11, 2026 06:21

kshitijk4poor mentioned this pull request May 11, 2026

follow-up: hermes models status --probe (reachability check, extracted from #23369) #23614

Closed

kshitijk4poor force-pushed the feat/cli-inventory branch 2 times, most recently from 9643efc to ed624a4 Compare May 11, 2026 07:10

kshitijk4poor force-pushed the feat/cli-inventory branch from ed624a4 to 10d10ad Compare May 11, 2026 07:39

kshitijk4poor mentioned this pull request May 11, 2026

refactor(inventory): extract shared ConfigContext + build_models_payload #23666

Closed

kshitijk4poor closed this May 11, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat: hermes models / hermes providers — non-interactive inventory CLI (#23359)#23369

feat: hermes models / hermes providers — non-interactive inventory CLI (#23359)#23369
kshitijk4poor wants to merge 1 commit into
NousResearch:mainfrom
kshitijk4poor:feat/cli-inventory

kshitijk4poor commented May 10, 2026 •

edited

Loading

Uh oh!

kshitijk4poor commented May 11, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

kshitijk4poor commented May 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What this PR does

Why

What this PR is NOT doing

Modes / network behaviour

Public API exposed by hermes_cli/inventory.py

JSON contract (default scriptable shape)

Substrate gotchas the implementation respects

Tests

Lint diff

Files changed

How to verify

Out of scope (tracked separately)

Refs

Uh oh!

kshitijk4poor commented May 11, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

kshitijk4poor commented May 10, 2026 •

edited

Loading

Public API exposed by `hermes_cli/inventory.py`