viettranx eb6723d674 fix(pipeline): include tool-schema tokens in overhead + dynamic compact max_tokens ...

- Add TokenCounter.CountToolSchemas() to measure JSON schema size for all tools
- Include tool schemas in OverheadTokens calculation for accurate context usage
- Implement dynamic max_tokens: in/25 clamp [1024, 8192] for compaction
- Add characterization tests: count_tool_schemas_test.go
- Add overhead verification tests: context_stage_overhead_test.go, context_stage_tool_overhead_test.go
- Add integration tests: context_stage_integration_test.go
- Add compact tests: loop_compact_dynamic_max_test.go, loop_compact_max_tokens_test.go
- Add sanitize tests: loop_history_sanitize_max_tokens_test.go
- Add integration test: loop_compact_integration_test.go

2026-04-23 08:31:53 +07:00

..

dispatcher_audio_ctx_test.go

feat(tools): TTS voice/model resolver with agent-level precedence

2026-04-15 11:24:57 +07:00

evolution_guardrails_test.go

fix(evolution): fix 7 bugs in evolution flow — guardrails, cron, dedup, dead types

2026-04-10 10:13:58 +07:00

evolution_guardrails.go

fix(evolution): fix 7 bugs in evolution flow — guardrails, cron, dedup, dead types

2026-04-10 10:13:58 +07:00

extractive_memory_test.go

test(tasks,agent): cover TaskTicker and agent helpers

2026-04-11 21:22:23 +07:00

extractive_memory.go

feat(providers): support Codex OAuth pools with inherited routing defaults

2026-03-27 09:35:57 +07:00

image_gen_gate_test.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

inject_and_misc_test.go

test(tasks,agent): cover TaskTicker and agent helpers

2026-04-11 21:22:23 +07:00

inject.go

refactor: replace magic numbers and hardcoded strings with named constants (#269)

2026-03-19 13:59:32 +07:00

input_guard_test.go

Initial commit: GoClaw AI agent gateway

2026-02-22 14:58:07 +07:00

input_guard.go

Initial commit: GoClaw AI agent gateway

2026-02-22 14:58:07 +07:00

intent_classify_test.go

feat(chat): auto-abort agent on exact "stop" keyword during active run

2026-04-16 20:32:37 +07:00

intent_classify.go

feat(chat): auto-abort agent on exact "stop" keyword during active run

2026-04-16 20:32:37 +07:00

loop_bench_test.go

chore: apply go fix syntax upgrades (min/max, switch patterns)

2026-04-15 11:24:57 +07:00

loop_cancel_test.go

feat(infra): tracing recovery, browser cleanup, CLI fixes, UI workspace split (#709)

2026-04-05 21:32:59 +07:00

loop_compact_dynamic_max_test.go

fix(pipeline): include tool-schema tokens in overhead + dynamic compact max_tokens

2026-04-23 08:31:53 +07:00

loop_compact_integration_test.go

fix(pipeline): include tool-schema tokens in overhead + dynamic compact max_tokens

2026-04-23 08:31:53 +07:00

loop_compact_max_tokens_test.go

fix(pipeline): include tool-schema tokens in overhead + dynamic compact max_tokens

2026-04-23 08:31:53 +07:00

loop_compact.go

fix(pipeline): include tool-schema tokens in overhead + dynamic compact max_tokens

2026-04-23 08:31:53 +07:00

loop_context.go

feat: propagate RBAC role through dispatch + bypass file-writer grant for admins

2026-04-16 14:17:48 +07:00

loop_finalize.go

fix(agent): extract MEDIA: paths from LLM response before sanitize

2026-04-18 15:34:46 +07:00

loop_history_edge_test.go

fix(agent): preserve MediaRefs when merging consecutive same-role messages

2026-04-02 22:58:35 +07:00

loop_history_sanitize_max_tokens_test.go

fix(pipeline): include tool-schema tokens in overhead + dynamic compact max_tokens

2026-04-23 08:31:53 +07:00

loop_history_sanitize.go

fix(pipeline): include tool-schema tokens in overhead + dynamic compact max_tokens

2026-04-23 08:31:53 +07:00

loop_history_skills.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

loop_history_supplement.go

feat(channels): shared WriterLabel helper with prompt-injection sanitization

2026-04-18 22:49:51 +07:00

loop_history_test.go

feat(pruning)!: faithful port of TS context pruning + backfill migration

2026-04-15 11:24:57 +07:00

loop_history_toolnames.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

loop_history.go

feat(agent): inject <current_reply_target> into system prompt

2026-04-18 17:45:14 +07:00

loop_input_media.go

feat(trace): reliable stop/abort with ctx-aware streams and 2-phase router

2026-04-14 18:28:31 +07:00

loop_lazy_mcp_test.go

fix(mcp): unified grant revocation with per-agent tool group isolation

2026-04-16 17:07:56 +07:00

loop_mcp_user.go

fix(mcp): unified grant revocation with per-agent tool group isolation

2026-04-16 17:07:56 +07:00

loop_media_dedup.go

fix(files): heal legacy media URLs and deduplicate file references

2026-03-25 12:37:49 +07:00

loop_media_test.go

fix(agent): close symlink escape + path-alias dedup + TTS mime type

2026-04-18 15:40:24 +07:00

loop_media.go

fix(agent): revert path-normalizing dedup and symlink-resolved storage

2026-04-18 16:19:23 +07:00

loop_pipeline_adapter.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

loop_pipeline_callbacks.go

feat(agent): inject <current_reply_target> into system prompt

2026-04-18 17:45:14 +07:00

loop_pipeline_tool_callbacks_test.go

fix(agent): propagate SenderID in AgentEvent emitters

2026-04-17 20:45:48 +07:00

loop_pipeline_tool_callbacks.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

loop_run.go

fix(agent): propagate SenderID in AgentEvent emitters

2026-04-17 20:45:48 +07:00

loop_team_reminders.go

fix(agent): group session unresponsive during team task execution (#266)

2026-03-30 15:20:01 +07:00

loop_tool_filter.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

loop_tools.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

loop_tracing.go

feat(tts): per-tenant TTS config and per-agent voice override

2026-04-16 21:26:45 +07:00

loop_truncation_test.go

fix: preserve FinishReason on truncated tool calls + ParseError propagation (#605)

2026-03-31 14:33:33 +07:00

loop_types.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

loop_utils_test.go

fix(security): prevent cross-group session data leak in cron jobs

2026-04-13 11:04:19 +07:00

loop_utils.go

fix(security): prevent cross-group session data leak in cron jobs

2026-04-13 11:04:19 +07:00

loop_v3_force_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

loop.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

media_assistant_images_test.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

media_filename_test.go

feat(trace): reliable stop/abort with ctx-aware streams and 2-phase router

2026-04-14 18:28:31 +07:00

media_filename.go

feat(trace): reliable stop/abort with ctx-aware streams and 2-phase router

2026-04-14 18:28:31 +07:00

media_persist_test.go

feat(trace): reliable stop/abort with ctx-aware streams and 2-phase router

2026-04-14 18:28:31 +07:00

media_sanitize.go

fix sending image via lark (#587)

2026-03-30 22:15:34 +07:00

media_test.go

fix(agent): use forward-scan replaceFirstMediaTag for correct multi-ref ordering

2026-03-31 19:26:40 +07:00

media_tool_routing.go

feat(trace): reliable stop/abort with ctx-aware streams and 2-phase router

2026-04-14 18:28:31 +07:00

media.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

memoryflush_test.go

refactor(comments): remove plan/phase refs from agent identity hardening code

2026-04-11 21:22:23 +07:00

memoryflush.go

refactor(comments): remove plan/phase refs from agent identity hardening code

2026-04-11 21:22:23 +07:00

orchestration_mode_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

orchestration_mode.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

png_metadata_test.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

png_metadata.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

preview_prompt_test_helpers_test.go

feat(vault): tenant-wide rescan with nullable agent_id + media preview

2026-04-10 16:32:54 +07:00

preview_prompt_test.go

feat(vault): tenant-wide rescan with nullable agent_id + media preview

2026-04-10 16:32:54 +07:00

preview_prompt.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

prompt_builder_impl.go

feat(agent): inject <current_reply_target> into system prompt

2026-04-18 17:45:14 +07:00

prompt_builder_orchestration_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

prompt_config_types.go

feat(agent): inject <current_reply_target> into system prompt

2026-04-18 17:45:14 +07:00

pruning_integration_test.go

feat(pruning)!: faithful port of TS context pruning + backfill migration

2026-04-15 11:24:57 +07:00

pruning_test.go

feat(pruning): enable context pruning by default with cache-ttl mode

2026-04-19 21:00:13 +07:00

pruning.go

feat(pruning): enable context pruning by default with cache-ttl mode

2026-04-19 21:00:13 +07:00

resolver_helpers.go

fix(agent): smart delegation prompt + compaction pending state + block team tools in subagents

2026-03-31 11:44:54 +07:00

resolver.go

feat: native image_generation for Codex + OpenAI-compat providers (#1002)

2026-04-23 08:22:39 +07:00

retriever.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

router_abort_test.go

fix(backup): harden tenant restore preview and lookup handling (#920)

2026-04-16 15:49:50 +07:00

router_cache_canonicalize_test.go

feat(tts): per-tenant TTS config and per-agent voice override

2026-04-16 21:26:45 +07:00

router_invalidate_collision_test.go

refactor(comments): remove plan/phase refs from agent identity hardening code

2026-04-11 21:22:23 +07:00

router_invalidate_tenant_test.go

fix(cache): tenant-aware invalidation for builtin tools and skills

2026-04-11 23:22:17 +07:00

router_invalidate_test.go

fix(router): canonicalize cache key, exact-segment match, tenant-aware IsRunning

2026-04-11 21:22:23 +07:00

router_isrunning_test.go

refactor(comments): remove plan/phase refs from agent identity hardening code

2026-04-11 21:22:23 +07:00

router_tenant_subscriber_branch_test.go

fix(cache): tenant-aware invalidation for builtin tools and skills

2026-04-11 23:22:17 +07:00

router.go

feat(trace): reliable stop/abort with ctx-aware streams and 2-phase router

2026-04-14 18:28:31 +07:00

sanitize_config_leak_test.go

fix(security): disable config leak detection to prevent false positives

2026-03-10 23:11:02 +07:00

sanitize_directives_test.go

fix(agent): strip internal routing tags from outbound delivery (#416)

2026-03-24 10:24:10 +07:00

sanitize_silent_reply_test.go

fix(agent): broaden NO_REPLY detection to cover decorative variants

2026-04-18 21:46:26 +07:00

sanitize_thinking_test.go

fix(openai): Together-compatible requests with reasoning/stream/vision gating

2026-04-04 21:17:47 +07:00

sanitize.go

fix(agent): broaden NO_REPLY detection to cover decorative variants

2026-04-18 21:46:26 +07:00

skill_draft_template_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

skill_draft_template.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

suggestion_engine.go

fix(evolution): fix 7 bugs in evolution flow — guardrails, cron, dedup, dead types

2026-04-10 10:13:58 +07:00

suggestion_rules_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

suggestion_rules.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

systemprompt_bootstrap_test.go

test: failing tests for softened predefined bootstrap prompt

2026-04-15 12:41:00 +07:00

systemprompt_cache_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

systemprompt_mode_test.go

fix(vault): prevent vault_read id-namespace collision (#959)

2026-04-19 15:00:07 +07:00

systemprompt_provider_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

systemprompt_sections.go

feat(agent): inject TTS [[tts]] directive guidance when auto mode is "tagged"

2026-04-16 18:55:43 +07:00

systemprompt_self_evolve_test.go

feat(v3): core architecture redesign — pipeline, memory, vault, evolution, providers, orchestration (#790)

2026-04-09 21:15:19 +07:00

systemprompt_target_test.go

test(tools,agent): cover cross-target guard scenarios

2026-04-18 17:45:28 +07:00

systemprompt_team_test.go

fix(agent): reduce system prompt size and fix team context injection (#613)

2026-04-02 18:32:43 +07:00

systemprompt.go

feat(tools): add send_file for delivering existing workspace files

2026-04-23 08:31:53 +07:00

title_generate.go

fix(agent/title): disable thinking and raise max_tokens for Gemini

2026-04-15 13:37:37 +07:00

tool_result_truncation.go

feat: cap tool output at source + improve context pruning pipeline

2026-03-31 08:35:06 +07:00

tool_timing.go

feat(agent): adaptive tool timing with slow tool notification

2026-03-19 13:35:57 +07:00

toolloop_test.go

fix(agent): uniqueness-aware loop detector + auto-fail killed team tasks (#506)

2026-03-28 14:53:43 +07:00

toolloop.go

fix(agent): uniqueness-aware loop detector + auto-fail killed team tasks (#506)

2026-03-28 14:53:43 +07:00

types.go

feat(tts): per-tenant TTS config and per-agent voice override

2026-04-16 21:26:45 +07:00

user_identity_resolver_test.go

feat(agent): centralized tenant user identity resolution for credentials

2026-04-08 14:47:39 +07:00

user_identity_resolver.go

feat(agent): centralized tenant user identity resolution for credentials

2026-04-08 14:47:39 +07:00

workspace_sharing_test.go

feat: decouple memory/KG sharing from workspace folder sharing

2026-03-12 18:26:40 +07:00