| commit |
1.00 |
fix(ci): lower Linux release glibc baseline (#707) |
|
Commit message contains explicit AI assi |
2026-07-10 |
| commit |
1.00 |
fix(logging): keep verbose openai.agents DEBUG off sandbox s |
|
Commit message contains explicit AI assi |
2026-07-10 |
| commit |
1.00 |
fix(runtime): swallow torn-down docker socket in sandbox del |
|
Commit message contains explicit AI assi |
2026-07-10 |
| commit |
1.00 |
fix(providers): declare bedrock + vertex extras and add prov |
|
Commit message contains explicit AI assi |
2026-07-07 |
| commit |
1.00 |
feat(report): tag SARIF rules with STRIDE legs derived from |
|
Commit message contains explicit AI assi |
2026-07-07 |
| commit |
1.00 |
feat(report): SARIF 2.1.0 emitter for CI / code-scanning int |
|
Commit message contains explicit AI assi |
2026-07-03 |
| commit |
1.00 |
test: add unit tests for config loader (strix/config/loader. |
|
Commit message contains explicit AI assi |
2026-06-30 |
| pr |
0.30 |
feat: implement retry logic for restartable stream errors in |
|
Somewhat formal summary, but technical l |
2026-07-01 |
| pr |
0.20 |
fix: recover from BadRequestError caused by oversized payloa |
|
Somewhat formal summary but strong domai |
2026-04-20 |
| pr |
0.20 |
Feature/configurable memory compression |
|
Slight formal tone, but technical conten |
2026-04-16 |
| pr |
0.20 |
feat(skills): expand llm_prompt_injection with Model DoS and |
|
Slightly more formal but highly technica |
2026-07-11 |
| pr |
0.20 |
added disposable email skill |
|
Slightly formal, but use-case focus with |
2026-06-11 |
| pr |
0.10 |
docs(reporting): add CVSS calibration guidance to reduce sev |
|
Uses domain-specific examples, concise e |
2026-07-21 |
| pr |
0.10 |
fix(bedrock): give Claude an explicit, clamped output ceilin |
|
Highly technical with error context and |
2026-07-02 |
| pr |
0.10 |
fix: retry transient 'thinking blocks cannot be modified' 40 |
|
Technical, cites precise error, no AI-li |
2026-04-27 |
| pr |
0.10 |
feat(dedupe): add dedicated deduplication model |
|
Cites motivation and details, mostly hum |
2026-07-21 |
| pr |
0.10 |
feat(verification): independent adversarial finding verifica |
|
Domain-specific and concise with clear c |
2026-07-21 |
| pr |
0.10 |
Fine-grained model control: skill routing, budget fallbacks, |
|
Contains jargon and detailed config with |
2026-07-17 |
| pr |
0.10 |
fix(core): Graceful handling of NFE loops and incremental SA |
|
Good technical detail but standard tone |
2026-07-19 |
| pr |
0.10 |
fix: tolerate string-encoded skills param for LLM provider c |
|
Free text uses domain context and common |
2026-07-15 |
| pr |
0.10 |
feat(llm): enable Bedrock/Anthropic prompt caching for Claud |
|
Detailed, domain-specific PR explanation |
2026-07-15 |
| pr |
0.10 |
feat(cli): --run-name for deterministic, resumable run dirs |
|
Explains with concrete use cases, domain |
2026-07-16 |
| pr |
0.10 |
feat(tools): batch_view_files / batch_terminal_execute — san |
|
Natural technical writing, contextual ex |
2026-07-16 |
| pr |
0.10 |
fix(runtime): close the docker client on session cleanup |
|
Technical and diagnostic, code and error |
2026-07-16 |
| pr |
0.10 |
fix(runtime): cap sandbox container logs to prevent host dis |
|
Contains domain jargon, technical detail |
2026-07-16 |
| pr |
0.10 |
feat(runtime): session-setup hook (register_session_setup) |
|
Technical context, code snippets, and di |
2026-07-16 |
| pr |
0.10 |
feat(config): accept GROQ_API_KEY for Groq-hosted models |
|
Uses domain-specific terms, structured b |
2026-07-03 |
| pr |
0.10 |
fix: skip Unix permission test on Windows (#648) |
|
Brief, technical, with platform-specific |
2026-07-03 |
| pr |
0.10 |
fix(agents): force tty for exec_command on Windows |
|
In-depth, technical, with incomplete sen |
2026-07-11 |
| pr |
0.10 |
expand docs folder README for contributors |
|
Expands content with practical contribut |
2026-07-02 |
| pr |
0.10 |
fix(proxy): recompute Content-Length when repeat_request rep |
|
Technical issue+solution structure, conc |
2026-06-29 |
| pr |
0.10 |
test: cover report helpers _parse_dedupe_response and litell |
|
Dense technical test details, abbreviati |
2026-06-26 |
| pr |
0.10 |
fix: declare socksio and guard tracing init against SOCKS-pr |
|
Contains precise technical details and i |
2026-06-26 |
| pr |
0.10 |
fix(interface): correct copy-diff classification and workspa |
|
Dense domain-specific explanation, natur |
2026-06-26 |
| pr |
0.10 |
fix(core): return True when image-strip retry succeeds in st |
|
Explains logic change and bug clearly in |
2026-06-26 |
| pr |
0.10 |
fix(interface): exit cleanly on unhandled LLM BadRequestErro |
|
Technical problem/solution, concise and |
2026-06-26 |
| pr |
0.10 |
Migrate task runner from Makefile to mise |
|
Structured PR but with domain-specific r |
2026-06-19 |
| pr |
0.10 |
fix(proxy): actionable errors for bad list_requests/list_sit |
|
Explains bug and fix in clear, domain-sp |
2026-06-15 |
| pr |
0.10 |
Fix Bedrock reasoning for adaptive-thinking models (Claude O |
|
Diagnostic and technical with error cont |
2026-06-09 |
| pr |
0.10 |
Feat/secure credentials |
|
Direct, bullet-based style; high technic |
2026-06-09 |
| pr |
0.10 |
Honor LLM_TIMEOUT, expose generation params, and harden blac |
|
Concise, domain-specific, natural senten |
2026-06-09 |
| pr |
0.10 |
Add Podman as working runtime backend |
|
Uses domain jargon, no ChatGPT-style phr |
2026-05-28 |
| pr |
0.10 |
Fix memory compressor timeout fallback |
|
Terse, technical, specific, no formaliti |
2026-05-26 |
| pr |
0.10 |
perf(tui): fix CPU lockup and O(N) scans during waiting/runn |
|
Domain-specific, informal tone, no AI si |
2026-05-22 |
| pr |
0.10 |
[codex] clarify Vertex AI env vars |
|
Clear technical summary, direct, and inf |
2026-05-21 |
| pr |
0.05 |
feat(routing): per-agent models and skill-based child routin |
|
Technical, concise with some abbreviatio |
2026-07-21 |
| pr |
0.05 |
fix(tui): don't double-render finish_scan section headings |
|
Terse and diagnostic; uses code refs and |
2026-07-17 |
| pr |
0.05 |
feat: add opt-in completion nudge |
|
Domain-specific context, slightly inform |
2026-07-10 |
| pr |
0.05 |
Fix ordered-list spacing in TUI markdown renderer |
|
Brief, informal summary bullets, closes |
2026-07-10 |
| pr |
0.05 |
fix(interface): clarify cached vs new input token breakdown |
|
Technical, clear bullet structure, casua |
2026-07-05 |
| pr |
0.05 |
fix: prevent premature subagent termination, add attack plan |
|
Uses domain-specific language, error ref |
2026-07-06 |
| commit |
0.00 |
chore: release v1.2.0 |
|
Standard release commit, no AI signals. |
2026-07-21 |
| commit |
0.00 |
Local run viewer: email reports, run history, and the platfo |
|
Concise, technical summary with no AI ha |
2026-07-21 |
| commit |
0.00 |
docs(skills): add Active Directory / Kerberos domain testing |
|
Detailed, domain-specific commit with ge |
2026-07-21 |
| commit |
0.00 |
docs(reporting): add CVSS calibration guidance to reduce sev |
|
Domain-specific guidance and typical hum |
2026-07-21 |
| commit |
0.00 |
fix(report): prevent code-fence breakout in vulnerability ma |
|
Deep technical details, human phrasing, |
2026-07-21 |
| commit |
0.00 |
strip transfer encoding (#820) |
|
Brief/tacit commit, minimal detail, clea |
2026-07-21 |
| commit |
0.00 |
fix(proxy): recompute Content-Length when replaying a modifi |
|
Contains a co-author GitHub trailer, but |
2026-07-21 |
| commit |
0.00 |
docs(skills): add grafana_prometheus observability pivot ski |
|
Commit message uses concise, domain-spec |
2026-07-20 |
| commit |
0.00 |
recon asset discovery skill (#809) |
|
Commit messages are terse and domain-spe |
2026-07-19 |
| commit |
0.00 |
chore(llm): shorten timeout helper docstring; update tests |
|
Concise, domain-specific commit message; |
2026-07-18 |
| commit |
0.00 |
fix(llm): use a JSON-serializable per-turn model timeout |
|
Technical language, references to code/d |
2026-07-18 |
| commit |
0.00 |
fix(llm): use httpx.Timeout read-inactivity for per-turn mod |
|
Direct summary with technical phrasing; |
2026-07-18 |
| commit |
0.00 |
fix(llm): add per-turn model request timeout so stalled stre |
|
Terse, technical message; no AI indicato |
2026-07-18 |
| commit |
0.00 |
fix(llm): retry statusless mid-stream provider errors (quota |
|
Domain-specific, concise, references to |
2026-07-17 |
| commit |
0.00 |
fix(llm): retry statusless mid-stream provider errors (quota |
|
Similar to prior; clear technical focus, |
2026-07-17 |
| commit |
0.00 |
fix(report): restore cost tracking for OpenRouter and other |
|
— |
2026-07-17 |
| commit |
0.00 |
revert(proxy): drop overfit Caido reconnect/HTTPQL band-aids |
|
— |
2026-07-17 |
| commit |
0.00 |
fix(proxy,tooling): serialize+reconnect Caido client, action |
|
— |
2026-07-17 |
| commit |
0.00 |
fix(prompt): down-rate or skip findings on demo data / demo |
|
— |
2026-07-17 |
| commit |
0.00 |
fix(runtime): close the docker client on session cleanup (#7 |
|
Standard commit message style; brief and |
2026-07-16 |
| commit |
0.00 |
fix(runtime): cap sandbox container logs to prevent host dis |
|
Domain-specific commit message; no AI st |
2026-07-16 |
| commit |
0.00 |
docs(python skill): recommend a task-unique PoC filename to |
|
Direct, human-written doc change; concis |
2026-07-16 |
| commit |
0.00 |
docs(python skill): use a distinctive PoC filename to avoid |
|
Clear domain context, no AI-typical phra |
2026-07-16 |
| commit |
0.00 |
docs(prompts,skills): stop hardcoding /workspace/scratch pat |
|
Technical explanation, typical human com |
2026-07-16 |
| commit |
0.00 |
docs(tools): guide proportional wait_for_message timeouts (# |
|
— |
2026-07-16 |
| commit |
0.00 |
docs(skills): add research-backed katana output-reduction fl |
|
Uses specific tool flags and domain term |
2026-07-16 |
| commit |
0.00 |
docs(prompts,skills): scope cleanup to own files; dedupe JSO |
|
Informal, references specific review and |
2026-07-16 |
| commit |
0.00 |
docs(prompts,skills): bound recon output for shared-disk hyg |
|
Detailed, project-specific, uses domain |
2026-07-16 |
| commit |
0.00 |
Revert "fix(runtime): retry transient sandbox startup failur |
|
— |
2026-07-16 |
| commit |
0.00 |
fix(runtime): bound nano_cpus to docker's int64 NanoCPUs ran |
|
Technical, concise, updates bound for do |
2026-07-16 |
| commit |
0.00 |
fix(runtime): also suppress OverflowError for non-finite STR |
|
Concise technical fix; domain context pr |
2026-07-16 |
| commit |
0.00 |
feat(runtime): opt-in resource limits for docker sandbox con |
|
Technical commit message with abbreviati |
2026-07-16 |
| commit |
0.00 |
fix(core): bound per-agent image memory (proactive budget + |
|
— |
2026-07-16 |
| commit |
0.00 |
feat(runtime): resolve sandbox ports over a shared Docker ne |
|
— |
2026-07-15 |
| commit |
0.00 |
fix(runtime): retry transient sandbox startup failures (#768 |
|
Commit messages are terse with technical |
2026-07-15 |
| commit |
0.00 |
fix(runtime): stage symlink-safe copies for LocalDir uploads |
|
Technical language, concise commits, dom |
2026-07-14 |
| commit |
0.00 |
chore: release v1.1.0 (#765) |
|
— |
2026-07-14 |
| commit |
0.00 |
Default sandbox exec commands to Bash (#764) |
|
Standard Git commit, human co-author, te |
2026-07-14 |
| commit |
0.00 |
Update README |
|
Extremely brief update message, no AI si |
2026-07-14 |
| commit |
0.00 |
Warn when configured LLM is not frontier-recommended (#586) |
|
— |
2026-07-14 |
| commit |
0.00 |
chore(telemetry): minor telemetry updates (#761) |
|
— |
2026-07-13 |
| commit |
0.00 |
feat(llm): change OpenRouter LLM request headers (#760) |
|
— |
2026-07-13 |
| commit |
0.00 |
fix(tui): restore snappy sweep/progress animation frame rate |
|
— |
2026-07-13 |
| commit |
0.00 |
feat: add weak password detection skill (#621) (#654) |
|
Standard conventional commit, no AI hall |
2026-07-13 |
| commit |
0.00 |
Deduplicate scan ended telemetry (#758) |
|
Commit messages and steps are natural an |
2026-07-13 |
| commit |
0.00 |
docs(prompts): strengthen report guidance (severity, chainin |
|
— |
2026-07-13 |
| commit |
0.00 |
fix(reporting): require advisory_cvss for dependency finding |
|
— |
2026-07-13 |
| commit |
0.00 |
Use core LiteLLM dependency (#752) |
|
Terse, domain-specific commit, no AI phr |
2026-07-12 |
| commit |
0.00 |
Add dependency reporting fields (#751) |
|
Straightforward commit message, typical |
2026-07-12 |
| commit |
0.00 |
fix(deps): cap openai<2.45 and add litellm[proxy] so fresh i |
|
Commit messages and trailers are human s |
2026-07-12 |
| commit |
0.00 |
Add root scan prompt options (#750) |
|
Simple commit message, fits human patter |
2026-07-12 |
| commit |
0.00 |
Add skill directory registration (#746) |
|
Human-like phrasing, no formal or AI pat |
2026-07-12 |
| commit |
0.00 |
Allow scan agent tool registration (#733) |
|
Terse commit, uses standard technical PR |
2026-07-12 |
| commit |
0.00 |
Support routed OpenAI required tool choice (#732) |
|
Concise PR title, normal routing/jargon, |
2026-07-10 |
| commit |
0.00 |
feat(settings): add force_required_tool_choice to LlmSetting |
|
Commit subject lines have domain terms a |
2026-07-10 |
| commit |
0.00 |
Address Greptile review: GCP and Auth0 recon guidance |
|
Human-written bullet points and technica |
2026-07-04 |
| commit |
0.00 |
Add GCP and Auth0 security skills |
|
Brief description, domain-specific terms |
2026-07-03 |
| commit |
0.00 |
fix(providers): match google submodule imports and walk full |
|
— |
2026-07-10 |
| commit |
0.00 |
test(providers): cover wrapped bedrock import errors |
|
Standard test-related commit message, hi |
2026-07-07 |
| commit |
0.00 |
fix(providers): show vertex extra hint for wrapped import er |
|
Very concise, specific fix message, no A |
2026-07-07 |
| commit |
0.00 |
fix(tui): key render cache by content string and return copi |
|
— |
2026-07-10 |
| commit |
0.00 |
fix(tui): reduce scroll stutter by throttling UI refresh and |
|
Informal tone, technical detail, natural |
2026-07-02 |
| commit |
0.00 |
fix(report): omit SARIF provenance for multiple repos (#726) |
|
Standard bugfix commit with SARIF jargon |
2026-07-10 |
| commit |
0.00 |
fix(container): allow configured Caido UI domains (#723) |
|
Short, technical fix commit, uses domain |
2026-07-10 |
| commit |
0.00 |
fix(session): use HTTPS scheme for Caido endpoint if TLS is |
|
Technical commit message with domain-spe |
2026-07-10 |
| commit |
0.00 |
Add target list CLI option (#711) |
|
Succinct commit message and changes; no |
2026-07-07 |
| commit |
0.00 |
test: add report writer artifact tests (#667) |
|
Concise test summary typical of human-wr |
2026-07-06 |
| commit |
0.00 |
fix(config): make env vars win over persisted JSON across al |
|
Detailed technical explanation with doma |
2026-07-06 |
| commit |
0.00 |
Update README (#705) |
|
Generic 'Update README' commit with no e |
2026-07-06 |
| commit |
0.00 |
Add LLM Prompt Injection skill (vulnerabilities) (#616) |
|
Short ticket title style; no signs of AI |
2026-07-06 |
| commit |
0.00 |
fix(tui): show 'more content available' for view_request ove |
|
Human-like concise technical fix with in |
2026-07-06 |
| commit |
0.00 |
Add five security skills: OAuth, AWS, prototype pollution, d |
|
Technical content, includes jargon and h |
2026-07-03 |
| commit |
0.00 |
fix: report cost for streamed OpenRouter calls (#634) |
|
Concise, informal technical phrasing; no |
2026-07-03 |
| commit |
0.00 |
fix: avoid note ID collisions (#630) |
|
Short, domain-specific fix phrasing; no |
2026-07-03 |
| commit |
0.00 |
fix grammer (#642) |
|
Standard commit message with typos and i |
2026-07-03 |
| commit |
0.00 |
fix(report): correct csv_path indentation in write_vulnerabi |
|
Technical, domain-specific commit; clear |
2026-07-02 |
| commit |
0.00 |
Fix non-atomic CSV and MD writes to prevent corruption on cr |
|
Brief, domain-specific commit message; n |
2026-07-02 |
| commit |
0.00 |
Remove collection of unhandled exception error messages from |
|
Concise summary typical of human commit |
2026-06-30 |
| commit |
0.00 |
chore(deps): refresh uv.lock to latest compatible versions ( |
|
Chore commit with domain wording; normal |
2026-06-30 |
| commit |
0.00 |
Update readme (#607) |
|
Very brief message, standard human pract |
2026-06-30 |
| commit |
0.00 |
Readme update |
|
Terse message for routine documentation |
2026-06-30 |
| commit |
0.00 |
fix: stop gracefully with resume hint on persistent RateLimi |
|
Technical commit with domain terminology |
2026-06-29 |
| commit |
0.00 |
fix(core): collapse child agent initial input into a single |
|
Domain-specific commit, uses jargon, not |
2026-06-29 |
| commit |
0.00 |
Support large target repos with with bind-mount option. (#57 |
|
Technical changelog with jargon and info |
2026-06-22 |
| commit |
0.00 |
Add configurable token / cost usage limits (#576) |
|
Detailed commit message with domain jarg |
2026-06-22 |
| commit |
0.00 |
fix: route ollama models through ollama_chat so tool calling |
|
Co-author trailer is a human email, mess |
2026-06-16 |
| commit |
0.00 |
Bump 1.0.3 -> 1.0.4 (#557) |
|
Version bump commit, minimal content, no |
2026-06-09 |
| commit |
0.00 |
Strip ANSI escapes and control bytes from terminal tool outp |
|
Concise message using domain-specific la |
2026-06-09 |
| commit |
0.00 |
Strip all images from session on vision-rejection, not just |
|
Brief, specific phrasing consistent with |
2026-06-09 |
| commit |
0.00 |
Swallow sandbox container races in the stream consumer (#552 |
|
Terse and technical, typical of develope |
2026-06-09 |
| commit |
0.00 |
Make TUI quit instant by SIGKILL-ing the sandbox container ( |
|
Informal, imperative tone with technical |
2026-06-09 |
| commit |
0.00 |
Bump 1.0.2 -> 1.0.3 (#537) |
|
Version bump commit, no AI hallmarks pre |
2026-06-09 |
| commit |
0.00 |
Simplify cost ledger to one bucket (#531) |
|
Very concise and technical, with no sign |
2026-06-08 |
| pr |
0.00 |
fix(prompt): make root agent orchestrate-only and fold fixin |
|
— |
2026-07-21 |
| pr |
0.00 |
fix(core): compile fallback report if agent misses lifecycle |
|
Template structure with authentic techni |
2026-07-03 |
| pr |
0.00 |
Add explicit CSPT detection skill, categorization, fixtures, |
|
Contains domain jargon and review respon |
2026-07-16 |
| pr |
0.00 |
sandbox: shrink image by ~2GB (go modcache, locales, docs) |
|
Detailed technical changelog, natural hu |
2026-05-04 |
| pr |
0.00 |
chore: release v1.2.0 |
|
— |
2026-07-21 |
| pr |
0.00 |
Local run viewer: email reports, run history, and the platfo |
|
Free-text is technical and direct; no AI |
2026-07-20 |
| pr |
0.00 |
docs(skills): add Active Directory / Kerberos domain testing |
|
Direct, domain-specific skill addition; |
2026-07-21 |
| pr |
0.00 |
feat(tui): add manual compact command |
|
Terse changelog, includes test output, n |
2026-07-07 |
| pr |
0.00 |
fix: strip chatgpt transcript tags from agent messages |
|
Direct language, specific details, no ge |
2026-07-10 |
| pr |
0.00 |
fix: propagate sandbox proxy environment aliases |
|
Brief, technical, includes error output |
2026-07-10 |
| pr |
0.00 |
Handle large exec_command output and surface scan errors as |
|
Casual, terse, domain-specific phrasing |
2026-06-22 |
| pr |
0.00 |
fix(report): prevent code-fence breakout in vulnerability ma |
|
Technical explanation with informal, pre |
2026-07-20 |
| pr |
0.00 |
strip transfer encoding |
|
Extremely terse and typical of human com |
2026-07-21 |
| pr |
0.00 |
fix(proxy): recompute Content-Length when replaying a modifi |
|
Uses domain jargon and casual but techni |
2026-07-20 |
| pr |
0.00 |
Close two follow-up viewer authz gaps (session-gate history |
|
— |
2026-07-20 |
| pr |
0.00 |
Harden local viewer authorization (session capability, histo |
|
— |
2026-07-20 |
| pr |
0.00 |
docs(skills): add grafana_prometheus observability pivot ski |
|
Concise, technical phrasing typical of h |
2026-07-20 |
| pr |
0.00 |
recon asset discovery skill |
|
Message is minimal, just a PR template t |
2026-07-19 |
| pr |
0.00 |
feat(cli): update notifications + self-update (strix --updat |
|
Brief, informal commit title with dev ja |
2026-07-19 |
| pr |
0.00 |
feat(cli): update notifications + self-update (strix --updat |
|
— |
2026-07-19 |
| pr |
0.00 |
fix(telemetry): stop finding() telemetry from blocking the a |
|
Filled template, uses code paths, error |
2026-07-04 |
| pr |
0.00 |
fix(llm): use a JSON-serializable per-turn model timeout |
|
— |
2026-07-18 |
| pr |
0.00 |
fix(llm): add per-turn model request timeout so stalled stre |
|
— |
2026-07-18 |
| pr |
0.00 |
fix(llm): retry statusless mid-stream provider errors (quota |
|
— |
2026-07-16 |
| pr |
0.00 |
Fall back to token-based cost estimation when observed LiteL |
|
— |
2026-07-17 |
| pr |
0.00 |
fix(report): restore cost tracking for OpenRouter and other |
|
— |
2026-07-17 |
| pr |
0.00 |
revert(proxy): drop overfit Caido reconnect/HTTPQL band-aids |
|
— |
2026-07-17 |
| pr |
0.00 |
fix(proxy,tooling): serialize+reconnect Caido client, action |
|
— |
2026-07-17 |
| pr |
0.00 |
fix(prompt): down-rate or skip findings on demo data / demo |
|
— |
2026-07-17 |
| pr |
0.00 |
Pulling |
|
Single informal word, clear human author |
2026-07-16 |
| pr |
0.00 |
docs(prompts,skills): stop hardcoding /workspace/scratch pat |
|
— |
2026-07-16 |
| pr |
0.00 |
docs(tools): guide proportional wait_for_message timeouts |
|
— |
2026-07-16 |
| pr |
0.00 |
docs(prompts,skills): bound recon output for shared-disk hyg |
|
— |
2026-07-16 |
| pr |
0.00 |
Revert "fix(runtime): retry transient sandbox startup failur |
|
— |
2026-07-16 |
| pr |
0.00 |
feat(runtime): opt-in resource limits for docker sandbox con |
|
— |
2026-07-16 |
| pr |
0.00 |
fix(core): bound per-agent image memory (proactive budget + |
|
— |
2026-07-15 |
| pr |
0.00 |
Need to get the latest updates |
|
Message is short, informal, and shows a |
2026-07-15 |
| pr |
0.00 |
feat(runtime): resolve sandbox ports over a shared Docker ne |
|
— |
2026-07-15 |
| pr |
0.00 |
feat(resume): opt-in grounded retraction of fixed findings |
|
Uses domain jargon and natural, context- |
2026-07-15 |
| pr |
0.00 |
Update vertex.mdx |
|
No free text; single generic template-ac |
2026-07-14 |
| pr |
0.00 |
fix: prevent context-window crash from huge tool outputs |
|
Terse, practical jargon and natural clos |
2026-07-11 |
| pr |
0.00 |
perf(runtime): import local sources via single tar put_archi |
|
Direct domain language, clear problem/so |
2026-07-11 |
| pr |
0.00 |
fix: reject submit-only XSS false positives |
|
Abbreviated, technical explanation using |
2026-07-12 |
| pr |
0.00 |
fix(tests): skip symlink test when Windows lacks privilege |
|
Direct technical explanation; specific t |
2026-07-11 |
| pr |
0.00 |
perf(runtime): import local sources via single tar put_archi |
|
Direct domain language, clear problem/so |
2026-07-05 |
| pr |
0.00 |
fix(runtime): retry transient sandbox startup failures |
|
No free-text content; only a terse templ |
2026-07-15 |
| pr |
0.00 |
fix(runtime): retry transient sandbox startup failures |
|
No free-text content; only a terse templ |
2026-07-15 |
| pr |
0.00 |
Warn when configured LLM is not frontier-recommended |
|
— |
2026-06-26 |
| pr |
0.00 |
fix(runtime): stage symlink-safe copies for LocalDir uploads |
|
Technical details, concise, clear monore |
2026-07-14 |
| pr |
0.00 |
chore: release v1.1.0 |
|
— |
2026-07-14 |
| pr |
0.00 |
Default sandbox exec commands to Bash |
|
Fills template; includes terse tech summ |
2026-07-14 |
| pr |
0.00 |
Update README |
|
No free-text content present, only templ |
2026-07-14 |
| pr |
0.00 |
feat: add support for setup script execution before scans |
|
Practical CLI documentation, abbreviatio |
2026-07-01 |
| pr |
0.00 |
fix: prevent long scans from hanging or dying on fd exhausti |
|
Detailed technical narrative, failure an |
2026-06-29 |
| pr |
0.00 |
fix(interface): drop dead reasoning-effort branches, fix sta |
|
Concise, practical discussion of code ch |
2026-06-26 |
| pr |
0.00 |
Testing |
|
Single word, lacks any AI hallmarks. |
2026-06-06 |
