hermes-agent/tools
teknium1 369075dc95 feat(tools): progressive tool disclosure for MCP and plugin tools
Adds Tool Search, a structured-tools progressive-disclosure layer that
replaces MCP and non-core plugin tools in the model-visible tools array
with three bridge tools (tool_search / tool_describe / tool_call) when
the deferrable surface would consume more than a configurable percentage
of the active model's context window. Core Hermes tools are never deferred.

Default mode is 'auto' with a 10% context threshold, so small toolsets
pay no overhead. Set tools.tool_search.enabled to 'on' to force or 'off'
to disable.

Design carefully reflects the OpenClaw production failure modes
documented in the openclaw-tool-search-report:

  - Core tools never defer (toolsets._HERMES_CORE_TOOLS). Addresses the
    'tools silently missing from isolated cron turns' regression class
    (openclaw#84141) by construction: there is no code path that can
    drop a core tool.
  - Catalog is stateless across turns — rebuilt from the live tool-defs
    list on every assembly. No session-keyed Map that can drift out of
    sync with the registry.
  - tool_call unwraps the bridge call before any hook fires, so plugin
    pre/post hooks, guardrails, approval flows, and the activity feed
    all see the underlying tool name, not the bridge (addresses
    openclaw#85588 and the verbose-mode complaint on openclaw#79823).
  - The unwrap happens in both the parallel and sequential paths of
    agent/tool_executor.py and also in handle_function_call, so direct
    callers (sandboxed code, eval harnesses) are covered too.
  - Bridge tools cannot invoke each other (recursion guard) and cannot
    invoke core tools (those must be called directly).
  - Tools mode only — no JS-sandbox code-mode. Keeps the surface small.
  - Token estimation via cheap char/4 heuristic; precision isn't needed
    for the threshold decision.

Files:
  - tools/tool_search.py — new module (BM25 retrieval, classification,
    threshold gate, bridge dispatch, unwrap helper).
  - tests/tools/test_tool_search.py — 35 tests including the OpenClaw
    #84141 regression guard.
  - model_tools.py — wires assembly into _compute_tool_definitions as the
    final step, adds skip_tool_search_assembly kwarg so the bridge can
    see the real catalog, dispatches the three bridge tools.
  - agent/tool_executor.py — unwraps tool_call in both parallel and
    sequential parsing loops so checkpointing, guardrails, plugin hooks,
    and tool-progress callbacks all observe the underlying tool name.
  - hermes_cli/config.py — DEFAULT_CONFIG['tools']['tool_search'] block.
  - website/docs/user-guide/features/tool-search.md — user docs.

Validation:
  - 35/35 new tests pass.
  - Existing tool/registry/model_tools/config/coercion/executor tests
    (82 + 74 + small adjacents) green.
  - Live E2E: 20 fake MCP tools registered, get_tool_definitions returns
    3 bridges, tool_search returns top 3 hits, tool_describe returns
    full schema, tool_call dispatches to the real underlying handler
    and the underlying result is what the model sees.
  - Reserved-name recursion guard verified live.
  - Core-tool refusal via tool_call verified live.
2026-05-29 02:04:12 -07:00
..
computer_use chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
environments fix(security): narrow Bedrock subprocess strip to inference bearer token only 2026-05-29 01:48:08 -07:00
neutts_samples refactor(tts): replace NeuTTS optional skill with built-in provider + setup flow 2026-03-17 02:33:12 -07:00
__init__.py Merge branch 'main' into rewbs/tool-use-charge-to-subscription 2026-03-31 08:48:54 +09:00
ansi_strip.py fix: strip ANSI at the source — clean terminal output before it reaches the model 2026-03-23 07:43:12 -07:00
approval.py approval: add docker restart/stop/kill to DANGEROUS_PATTERNS (#33438) 2026-05-29 13:26:54 +10:00
binary_extensions.py fix(tools): address PR review — remove _extract_raw_output, BudgetConfig everywhere, read_file hardening 2026-04-08 02:24:32 -07:00
browser_camofox.py fix(browser): rewrite Camofox Docker loopback URLs (#25541) 2026-05-29 15:43:55 +10:00
browser_camofox_state.py feat(browser): add persistent Camofox sessions and VNC URL discovery (salvage #4400) (#4419) 2026-04-01 04:18:50 -07:00
browser_cdp_tool.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
browser_dialog_tool.py feat: auto-launch Chromium-family browser for CDP 2026-05-19 22:34:05 -07:00
browser_supervisor.py fix(async): close unscheduled coroutines in all threadsafe bridges (#26584) 2026-05-15 14:00:01 -07:00
browser_tool.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
budget_config.py chore: remove Atropos RL environments and tinker-atropos integration (#26106) 2026-05-15 10:36:38 +05:30
checkpoint_manager.py chore: ruff auto-fix PLR6201 — tuple → set in membership tests (#23937) 2026-05-11 11:13:25 -07:00
clarify_gateway.py feat(gateway): wire clarify tool with inline keyboard buttons on Telegram (#24199) 2026-05-12 16:33:33 -07:00
clarify_tool.py refactor: add tool_error/tool_result helpers + read_raw_config, migrate 129 callsites 2026-04-07 13:36:38 -07:00
code_execution_tool.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
computer_use_tool.py feat(computer-use): cua-driver backend, universal any-model schema 2026-05-08 11:07:38 -07:00
credential_files.py remove Vercel AI Gateway and Vercel Sandbox (#33067) 2026-05-27 00:43:32 -07:00
cronjob_tools.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
debug_helpers.py refactor: codebase-wide lint cleanup — unused imports, dead code, and inefficient patterns (#5821) 2026-04-07 10:25:31 -07:00
delegate_tool.py fix(delegation): preserve configured_provider name when runtime returns 'custom' 2026-05-17 11:40:05 -07:00
discord_tool.py feat: add Discord message deletion action 2026-05-07 05:11:09 -07:00
env_passthrough.py harden(env_passthrough): apply GHSA-rhgp-j443-p4rf filter to config.yaml path (#27794) 2026-05-25 03:35:23 -07:00
env_probe.py feat(prompt): universal task-completion guidance + local Python toolchain probe (#34340) 2026-05-28 22:26:09 -07:00
fal_common.py refactor(image_gen): port FAL backend to plugins/image_gen/fal 2026-05-22 04:10:45 -07:00
feishu_doc_tool.py perf(cli): cut ~19s from 'hermes' cold start (skills cache + lazy Feishu + no Nous HTTP) (#22138) 2026-05-08 16:39:32 -07:00
feishu_drive_tool.py perf(cli): cut ~19s from 'hermes' cold start (skills cache + lazy Feishu + no Nous HTTP) (#22138) 2026-05-08 16:39:32 -07:00
file_operations.py remove Vercel AI Gateway and Vercel Sandbox (#33067) 2026-05-27 00:43:32 -07:00
file_state.py feat(delegate): cross-agent file state coordination for concurrent subagents (#13718) 2026-04-21 16:41:26 -07:00
file_tools.py remove Vercel AI Gateway and Vercel Sandbox (#33067) 2026-05-27 00:43:32 -07:00
fuzzy_match.py fix(patch): widen new_string \t/\r unescape to all match strategies (#33733) 2026-05-28 03:27:20 -07:00
homeassistant_tool.py fix: clean up description escaping, add string-data tests 2026-04-13 04:45:07 -07:00
image_generation_tool.py feat(auth) normalise the way in which we check whether a user has free/paid access to nous portal so we can expose behaviour and error messages accordingly. 2026-05-28 00:19:31 -07:00
interrupt.py fix(interrupt): propagate to concurrent-tool workers + opt-in debug trace (#11907) 2026-04-17 20:39:25 -07:00
kanban_tools.py fix(kanban): bridge worker runtime activity to board heartbeat (#31752) 2026-05-29 00:05:58 -07:00
lazy_deps.py remove Vercel AI Gateway and Vercel Sandbox (#33067) 2026-05-27 00:43:32 -07:00
managed_tool_gateway.py fix(tools): add debug logging for token refresh and tighten domain check 2026-04-02 12:40:03 +11:00
mcp_oauth.py feat(mcp-oauth): accept 'skip' at paste prompt to bypass auth without disabling server (#32069) 2026-05-25 05:37:30 -07:00
mcp_oauth_manager.py fix(mcp-oauth): persist OAuth server metadata across process restarts (#21226) 2026-05-07 05:35:33 -07:00
mcp_tool.py fix(mcp): resolve bare npx/npm/node against /usr/local/bin 2026-05-29 10:05:42 +10:00
memory_tool.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
microsoft_graph_auth.py feat(msgraph): add auth and client foundation 2026-05-08 09:27:26 -07:00
microsoft_graph_client.py fix(msgraph): stream download_to_file body instead of buffering 2026-05-08 09:27:26 -07:00
mixture_of_agents_tool.py chore: ruff auto-fix C401, C416, C408, PLR1722 (#23940) 2026-05-11 11:20:58 -07:00
neutts_synth.py fix(tts): document NeuTTS provider and align install guidance (#1903) 2026-03-18 02:55:30 -07:00
openrouter_client.py refactor: route ad-hoc LLM consumers through centralized provider router 2026-03-11 20:02:36 -07:00
osv_check.py chore: ruff auto-fix PLR6201 — tuple → set in membership tests (#23937) 2026-05-11 11:13:25 -07:00
patch_parser.py fix(lint): skip per-file shell linter when LSP will handle the file (#29054) 2026-05-20 01:46:40 -05:00
path_security.py refactor: extract shared helpers to deduplicate repeated code patterns (#7917) 2026-04-11 13:59:52 -07:00
process_registry.py feat(cli): show live background terminal-process count in status bar (#32061) 2026-05-25 05:35:02 -07:00
registry.py security: sanitize tool error strings before injecting into model context (#26823) 2026-05-16 00:57:39 -07:00
schema_sanitizer.py fix(xai-responses): strip enum values containing '/' from tool schemas 2026-05-18 10:37:35 -07:00
send_message_tool.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
session_search_tool.py feat(session_search): single-shape tool with discovery, scroll, browse — no LLM (#27590) 2026-05-17 23:28:45 -07:00
skill_manager_tool.py fix(profiles): cross-profile soft guard on file-write tools + system-prompt hint (#31290) 2026-05-24 00:38:17 -07:00
skill_provenance.py fix(curator): only mark agent-created for background-review sediment (#19621) 2026-05-04 02:42:16 -07:00
skill_usage.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
skills_ast_audit.py refactor(skills): slim AST diagnostic to single entry point 2026-05-23 17:47:26 -07:00
skills_guard.py Revert "feat(skills): integrate NVIDIA/skills as a trusted skills hub tap" 2026-05-28 20:39:39 -07:00
skills_hub.py Revert "feat(skills): integrate NVIDIA/skills as a trusted skills hub tap" 2026-05-28 20:39:39 -07:00
skills_sync.py fix(skills): atomic lock write + drop dead _validate_category_name 2026-05-27 13:39:58 -07:00
skills_tool.py remove Vercel AI Gateway and Vercel Sandbox (#33067) 2026-05-27 00:43:32 -07:00
slash_confirm.py fix(async): close unscheduled coroutines in all threadsafe bridges (#26584) 2026-05-15 14:00:01 -07:00
terminal_tool.py fix(docker): cleanup_vm() default honors persist mode (don't kill container on session close) 2026-05-29 11:49:54 +10:00
threat_patterns.py feat(security): promptware defense — shared threat patterns + memory load-time scan + tool-result delimiters (#32269) 2026-05-25 14:52:24 -07:00
tirith_security.py fix(tirith): reject non-regular tar members during auto-install process 2026-05-28 02:49:26 -07:00
todo_tool.py chore: ruff auto-fix PLR6201 — tuple → set in membership tests (#23937) 2026-05-11 11:13:25 -07:00
tool_backend_helpers.py fix(tools): guard Path.home() against PermissionError in has_direct_modal_credentials (#33528) 2026-05-29 13:35:39 +10:00
tool_output_limits.py feat(skills): add design-md skill for Google's DESIGN.md spec (#14876) 2026-04-23 21:51:19 -07:00
tool_result_storage.py fix(tool-result-storage): persist via stdin to bypass 128 KB exec-arg cap (#22913) 2026-05-09 18:44:58 -07:00
tool_search.py feat(tools): progressive tool disclosure for MCP and plugin tools 2026-05-29 02:04:12 -07:00
transcription_tools.py feat(auth) normalise the way in which we check whether a user has free/paid access to nous portal so we can expose behaviour and error messages accordingly. 2026-05-28 00:19:31 -07:00
tts_tool.py feat(auth) normalise the way in which we check whether a user has free/paid access to nous portal so we can expose behaviour and error messages accordingly. 2026-05-28 00:19:31 -07:00
url_safety.py fix(url_safety): block IPv4-mapped IPv6 addresses to prevent SSRF bypass 2026-05-18 10:51:15 -07:00
video_generation_tool.py chore: ruff auto-fix PLR6201 resweep — tuple → set in membership tests (#27355) 2026-05-17 02:29:41 -07:00
vision_tools.py fix(vision): route auxiliary.vision.provider=openai to api.openai.com, skip text-only main (#31452) 2026-05-24 15:01:28 -07:00
voice_mode.py fix(voice): honor PIPEWIRE_REMOTE in PortAudio fallback checks (#33473) 2026-05-29 13:30:17 +10:00
web_tools.py fix: re-export pruned names for tests that mock.patch or from-import them 2026-05-28 22:26:25 -07:00
website_policy.py chore(web): remove web_crawl tool + provider crawl plumbing (#33824) 2026-05-28 04:52:42 -07:00
x_search_tool.py chore: prune unused imports and duplicate import redefinitions 2026-05-28 22:26:25 -07:00
xai_http.py feat(web): add xAI Web Search provider plugin 2026-05-19 19:27:34 -07:00
yuanbao_tools.py Fix unsafe gateway media path delivery 2026-05-23 01:40:35 -07:00