agent · hermes-agent · Tangled

The agent that grows with you

hermes-agent agent

35 files

at refactor/extract-codex-adapter

kshitijk4poor · refactor: extract codex_responses logic into dedicated adapter · 6mo ago

Refactor Terminal and AIAgent cleanup

anthropic_adapter.py

feat(providers): extend request_timeout_seconds to all client paths

auxiliary_client.py

fix(aux): add session_search extra_body and concurrency controls

bedrock_adapter.py

feat: native AWS Bedrock provider via Converse API

codex_responses_adapter.py

refactor: extract codex_responses logic into dedicated adapter

context_compressor.py

feat(compression): summaries now respect the conversation's language

context_engine.py

refactor: remove dead code — 1,784 lines across 77 files (#9180)

context_references.py

fix(agent): fall back when rg is blocked for @folder references

copilot_acp_client.py

fix: handle httpx.Timeout object in CopilotACPClient (#11058)

credential_pool.py

fix(codex): Hermes owns its own Codex auth; stop touching ~/.codex/auth.json (#12360)

fix(display): render <missing old_text> in memory previews instead of empty quotes (#12852)

error_classifier.py

fix(error_classifier): handle dict-typed message fields without crashing

gemini_cloudcode_adapter.py

fix(gemini): assign unique stream indices to parallel tool calls

gemini_native_adapter.py

fix(gemini): sanitize tool schemas for Google providers

gemini_schema.py

fix(gemini): sanitize tool schemas for Google providers

google_code_assist.py

fix(gemini-cli): surface MODEL_CAPACITY_EXHAUSTED cleanly + drop retired gemma-4-26b (#11833)

google_oauth.py

feat(gemini): add Google Gemini CLI OAuth provider via Cloud Code Assist (free + paid tiers) (#11270)

fix(insights): hide cache read/write and cost metrics from display (#11477)

manual_compression_feedback.py

fix(gateway): make manual compression feedback truthful

memory_manager.py

feat(honcho): context injection overhaul, 5-tool surface, cost safety, session isolation (#10619)

memory_provider.py

refactor(memory): drop on_session_reset — commit-only is enough

model_metadata.py

fix(gemini-cli): surface MODEL_CAPACITY_EXHAUSTED cleanly + drop retired gemma-4-26b (#11833)

fix(gemini): hide stale and low-TPM Google models

nous_rate_guard.py

fix: Nous Portal rate limit guard — prevent retry amplification (#10568)

prompt_builder.py

docs(memory): steer agents to save declarative facts, not instructions (#12665)

prompt_caching.py

fix(prompt-caching): skip top-level cache_control on role:tool for OpenRouter

rate_limit_tracker.py

refactor: remove dead code — 1,784 lines across 77 files (#9180)

fix(security): add JWT token and Discord mention redaction (#10547)

feat(agent): add jittered retry backoff

skill_commands.py

fix: use absolute skill_dir for external skills (#10313) (#10587)

feat(plugins): namespaced skill registration for plugin skill bundles

subdirectory_hints.py

fix(agent): catch PermissionError in subdirectory hint discovery

title_generator.py

fix: title_generator no longer logs as 'compression' task

Refactor Terminal and AIAgent cleanup

usage_pricing.py

feat: native AWS Bedrock provider via Converse API