fix: per-provider error resilience and /providers remove TUI command by KooshaPari · Pull Request #3685 · tailcallhq/forgecode

KooshaPari · 2026-07-15T02:35:20Z

SummaryTwo fixes addressing provider error handling in the forge TUI:

1. Per-provider error resilience in /models

The get_all_provider_models() function in crates/forge_app/src/app.rs previously used collect::<anyhow::Result<Vec<_>>>() which short-circuits on the first error from any provider. This caused /models (and forge list models) to fail entirely if one provider returned a 500, even when other providers were healthy.

Fix: The concurrent results are now separated into successes and failures. Individual provider failures are logged via tracing::warn!. The operation only returns an error when every configured provider fails — matching the existing doc comment behavior.

Closes: #3684

2. `/providers remove` TUI command

Added a new /providers remove command (AppCommand::ProviderRemove) that lets users interactively select and remove a configured provider from the REPL. This mirrors the existing /logout command behavior but uses the natural remove sub-command name under the providers namespace.

Changes

crates/forge_app/src/app.rs: Switched get_all_provider_models from fail-fast collect to graceful per-provider error handling with warning logs.
crates/forge_main/src/model.rs: Added ProviderRemove variant to AppCommand enum and its name() mapping.
crates/forge_main/src/ui.rs: Added handler for AppCommand::ProviderRemove in on_command.

- Bootstrap deny.toml with license allowlist + advisory ignores - Add license = MIT to workspace.package (was missing) - Add license.workspace = true to all 27 crate manifests - Ignore transitive unmaintained (bincode, yaml-rust, paste, rustls-pemfile) - Ignore transitive vulns (hickory-proto, rustls-webpki) via aws-sdk/reqwest Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Pin all action refs to immutable SHAs across workflow files: - checkout@v4 → @11bd71901bbe5b1630ceea73d27597364c9af683 - checkout@v6 → @de0fac2e4500dabe0009e67214ff5f5447ce83dd - setup-node@v4/v5, setup-python@v4/v5, setup-go@v5 - upload-artifact@v4/v7, download-artifact@v4 - cache@v3/v4, github-script@v7 - configure-pages@v5/v6, deploy-pages@v4/v5 - upload-pages-artifact@v3/v5, dependency-review-action@v4 Fixes version-tag normalization (add v4/v5 tags where missing). Fixes double-SHA corruption artifacts from prior patching rounds. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Reduce README from 1124 to 169 lines (-85%). Keep: project name, description, quickstart, usage examples, why forge, installation, community, documentation link. Add fork disclaimer pointing to upstream tailcallhq/forgecode. Preserve all upstream content via pointer comment. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Reduce README from 1124 to 169 lines (-85%). Keep: project name, description, quickstart, usage examples, why forge, installation, community, documentation link. Add fork disclaimer pointing to upstream tailcallhq/forgecode. Preserve all upstream content via pointer comment. Co-authored-by: Phenotype Agent <agent@phenotype.ai> Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Verified resolved upstream; advisory no longer triggers. Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

…nale - Add 4 ignore entries for bincode 1.x (RUSTSEC-2025-0141), paste (2024-0436), rustls-pemfile (2025-0134), yaml-rust (2024-0320) — all transitive via upstream forgecode workspace deps; resolution depends on upstream tailcallhq/forgecode bumps. - Pre-existing fork-specific RUSTSEC-2026-* ignores preserved. - cargo deny check advisories: PASS.

…nale (#8) - Add 4 ignore entries for bincode 1.x (RUSTSEC-2025-0141), paste (2024-0436), rustls-pemfile (2025-0134), yaml-rust (2024-0320) — all transitive via upstream forgecode workspace deps; resolution depends on upstream tailcallhq/forgecode bumps. - Pre-existing fork-specific RUSTSEC-2026-* ignores preserved. - cargo deny check advisories: PASS. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Root cause: crossterm's cursor position CSI query times out (2s) when multiple concurrent sessions are running or terminal is under load. Fix: - Add error::is_cursor_timeout_error() to detect cursor position errors - Add terminal::get_cursor_position_with_retry() with backoff - Suppress cursor errors during shutdown in Ui::shutdown() - Add comprehensive tests for cursor error detection Fixes session crashes where user sees: 'cursor position could not be read within a normal duration' 'Resource temporarily unavailable (os error 35)' Tested: 337 tests pass (333 existing + 4 new cursor error tests)

Add summarization feature with: - llm_summarizer: Async LLM-based summarization service - adaptive_eviction: Importance-based eviction strategies - metrics: Summarization metrics tracking - prefilter: Pre-summarization filtering - Updated compaction config and strategy Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

…p-go Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Restore the workspace members array by listing all crate directories present on disk, removing invalid glob patterns. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

- Add [bans] section with recommended warnings - Update GitHub workflow files (trufflehog, stale, labels, release)

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Preserve two research dossiers capturing the 2026-06-29 state-of-the-art analysis for agentic coding CLI tools and competitor landscape. These were at risk of loss in the uncommitted canonical checkout. Co-authored-by: Phenotype Agent <agent@phenotype.ai> Co-authored-by: ForgeCode <noreply@forgecode.dev>

…#68) PolicyEngine previously returned Permission::Confirm (prompt user) when no policies were configured or no rule matched an operation. This is a default-allow posture: any unconfigured operation could proceed after a single user confirmation, with no explicit allowlist required. forge_tracker previously enabled PostHog telemetry by default (opt-out via FORGE_TRACKER=false) and transmitted raw CLI args and Prompt event content (user-supplied text) in every telemetry event. Changes (Phenotype-org security hardening, audit issue #58): - PolicyEngine: return Permission::Deny (not Confirm) when the policy list is empty or when no rule matches the requested operation. An explicit allowlist is now required to permit any operation. - tracking_enabled(): invert default from opt-out to opt-in. FORGE_TRACKER must be explicitly set to 'true' to enable telemetry; absent or any other value disables it. - redact_event_value(): replace Prompt and Error event content with '<redacted>' before any collector.collect() call so user-typed text and stack traces are never sent to PostHog. - redact_args(): replace positional CLI arg values with '<redacted>' at cache time; flag names (--foo) are kept for diagnostics but inline =values are stripped. Regression tests added in both crates (failing-first verified manually): - test_policy_engine_default_deny_when_no_policies - test_policy_engine_default_deny_when_no_rule_matches - test_tracking_disabled_by_default_when_env_absent - test_tracking_enabled_only_when_explicitly_true - test_prompt_event_value_is_redacted - test_error_event_value_is_redacted - test_non_sensitive_event_value_is_not_redacted - test_redact_args_replaces_positional_values - test_redact_args_keeps_flag_names_strips_inline_values Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Conv-list used SELECT * loading multi-MB context blobs for all 2585 convos. Lite metadata-only query + lazy context load. - ConversationSummary domain struct (metadata only, no context blob) - ConversationRecordLite Diesel struct with .select() on (id,title, created_at,updated_at,message_count,cwd,parent_id,workspace_id) - get_parent_conversations_lite() on all layers (repo→service→API) - ConversationSelector::select_conversation now returns ConversationId (callers lazy-load full Conversation only on open) - Users of get_parent_conversations() for list views switched to lite See docs/sessions/20260629-forgedev-ram/RAM_INVESTIGATION.md Co-authored-by: Phenotype Agent <agent@phenotype.ai> Co-authored-by: forge-dev <forge-dev@phenotype.ai>

Add comprehensive packaging specification for KooshaPari/forgecode fork. Covers Cargo binary rename (forge -> forge-dev), config directory rename (~/.forge -> ~/.forge-dev), shell plugin updates, per-OS installer formats (macOS .app, Windows MSI, Linux AppImage), universal CLI install script, side-by-side collision validation, and a 12-PR implementation plan. This is a spec-only PR with no code changes. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

…uency to Daily (#74) Profiling across 3 regimes showed the update-informer GitHub API call (update_informer::new(registry::GitHub, ...).check_version()) is synchronous and blocks every forge invocation for ~220ms when UpdateFrequency=Always (the previous default). Measured impact: forge info normal: ~300ms forge info (Never/agent): ~72ms M=32 parallel (with update): wall 1049ms -> throughput ~30 agents/s plateau M=64 parallel (with update): wall 2094ms -> throughput ~30 agents/s (flat) Two fixes: 1. Default UpdateFrequency from `Always` (0s interval = network on every run) to `Daily` (86400s interval; update_informer caches result, subsequent runs pay 0ms). Interactive users still get update prompts, just once per day. 2. is_non_interactive() guard in on_update(): skip the check entirely when stdin is not a TTY or CI/FORGE_NON_INTERACTIVE/FORGE_AGENT_MODE env vars are set. Agent fleets spawn hundreds of short-lived forge processes; the ~220ms per-spawn network tax is pure waste in that context. Remaining per-invocation cost breakdown (no update check): dyld + tokio + jemalloc init: ~47ms (irreducible without daemon pooling) SQLite pool open + PRAGMA: ~5-25ms Total non-LLM startup: ~72ms Regime 3 scaling plateau root cause (separate from update check): M=32: each forge proc spawns ~18 OS threads (tokio rt-multi-thread default) At M=32: 576 total OS threads -> macOS scheduler contention Throughput peaks at ~30-35 agents/s then falls super-linearly. Fix is a long-running daemon/pool pattern (future work; see Zig plan below). Zig hot-path candidates (not implemented this pass — ranked by impact): 1. Daemon socket: single forge process accepting tasks over Unix socket eliminates the ~47ms dyld+tokio-init overhead for all but the first spawn. Zig `posix_spawn` + kqueue IPC dispatcher is the right layer. 2. Spawn pool: pre-fork N forge workers, assign tasks via lock-free SPSC ring. Zig std.os.linux.io_uring or kqueue on macOS. This is the density multiplier — removes the 576-thread storm at M=32. 3. WAL checkpointer (already in Rust in checkpoint.rs) — stays Rust. SQLite open at M=32 costs only 53ms wall (scales fine); not the limit. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

…ilcallhq#2532) (#75) jemalloc on macOS ARM64 triggers upstream bug tailcallhq#2532: HPA probes every 4KB page, causing 14s cold init vs 25ms baseline (20x regression, measured PR#74). - Workspace Cargo.toml: move tikv-jemallocator to Linux-only; add mimalloc - forge_main/Cargo.toml: platform-conditional allocator deps - main.rs: cfg-gated #[global_allocator] (jemalloc on Linux, mimalloc on macOS) Co-authored-by: Phenotype Agent <agent@phenotype.ai>

… migration (#76) ~13,053 conversations with is_compressed=0 left ~2.9GB of plain JSON context on disk. New idempotent maintenance command compresses them in-place without modifying titles, timestamps, or any queryable columns. - ConversationRepository trait: add compress_uncompressed_contexts() - ConversationRepositoryImpl: raw SQL batch read + per-row zstd encode + UPDATE context_zstd / clear context / set is_compressed=1 - Wired through forge_repo -> forge_services -> forge_app -> forge_api - CLI: 'forge maintenance compress' subcommand with spinner + row counts - 2 unit tests: basic round-trip and idempotency Co-authored-by: Phenotype Agent <agent@phenotype.ai>

…sity (#77) Adds the forge-daemon density lever (forgecode#74 root cause: dyld+tokio init ~47ms×M + thread storm 18 OS threads×M = 576 threads at M=32). Architecture (Zig core + Rust orchestration, per feedback_zig_for_lowlevel): - forge-daemon/: Zig 0.16 project → two artifacts: libforge_daemon_core.a (C-ABI staticlib, consumed by Rust build.rs) forge-daemon binary (standalone Unix-socket daemon) Modules: kqueue_loop.zig (macOS kqueue, no EV_SET macro, std.c-based), socket.zig (Unix SOCK_STREAM, u32-LE length-prefix protocol), worker.zig (posix_spawn pool, nanosleep shutdown), ring.zig (lock-free SPSC task ring, RING_CAP=256), protocol.zig (JSON request/response parse), lib.zig (C-ABI exports: forge_daemon_start/stop/is_running/ socket_path/dispatch), main.zig (kqueue event loop, env-var config, multi-client) - crates/forge_daemon/: Rust orchestration (build.rs invokes zig build, links libforge_daemon_core.a). Exposes: DaemonDispatch::dispatch() — in-process C-ABI hot path (no socket) DaemonClient — async Unix-socket client (run/ping/shutdown) DaemonGuard — spawn/manage the standalone daemon binary - benchmarks/forge_daemon_bench/: M=8/16/32 daemon vs fork+exec benchmark. REAL BENCHMARK RESULTS (M=8/16/32, iters=5, /usr/bin/true, aarch64 macOS): M=8: fork+exec 3.1ms 2581 a/s → daemon 0.5ms 14904 a/s 5.77× faster M=16: fork+exec 5.4ms 2952 a/s → daemon 1.1ms 14763 a/s 5.00× faster M=32: fork+exec 9.2ms 3461 a/s → daemon 1.9ms 16497 a/s 4.77× faster The ~5× speedup is measured on process-launch overhead only (true binary, no LLM wait). Real forge workloads carry an additional ~47ms dyld+tokio init per spawn that the daemon eliminates entirely (measured in #74). The in-process DaemonDispatch path using posix_spawn via C-ABI also avoids tokio thread-storm (daemon uses one kqueue thread vs 18 OS threads per forge invocation). Tests: 2 Rust tests (daemon_not_running_initially, daemon_start_stop_lifecycle) + 1 Zig test (ring submit/consume round-trip) — all green. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Adds forge_repo_map crate — a tree-sitter-based module that walks a repository, parses source files with tree-sitter, and builds a compact symbol/structure map (files -> top-level functions, types, imports) that an agent can use as repo context. Supports Rust, TypeScript, JavaScript, Python, Go with tree-sitter grammars. gitignore-aware file walking via ignore crate. Builder API with max_files cap, configurable cwd, and stable text output format. Adds 19 tests covering: - Language detection (6 langs, fallback) - Rust: function, struct, enum, trait, type alias, impl, use import, unsafe - TypeScript: function, interface, class, type alias, enum, import - Python: function, class, import - Go: function, struct, interface, method, type alias, import - Builder: file discovery, gitignore respect, max_files, stable ordering, unsupported extension filtering, empty directory, text output format Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

…pool) (#78) Adds benchmarks/perf_harness which spawns forge (or a long-running placeholder) under three regimes (warmup=1, sustained=8, burst=32) and captures peak RSS, idle CPU, and proc stats into docs/perf/baseline-2026-07-02.json. macOS path uses /bin/ps; Linux path uses /proc/<pid>/{status,stat,fd}. Also fixes a pre-existing compile break in crates/forge_main/src/ui.rs:1044 where spinner.stop() was called with 0 args but the API takes Option<String>. This blocks building forge_main on toolchain 1.95, so the harness can't measure anything until it lands. Scorecard (commit 6f8f80a, macOS aarch64, tokio 10 workers): warmup workers= 1 forge_cold=9760 KiB mean_cpu=0.70% dur=406ms sustained workers= 8 mean_rss=1120 KiB mean_cpu=0.19% dur=549ms burst workers=32 mean_rss=1121 KiB mean_cpu=0.59% dur=582ms Co-authored-by: Phenotype Agent <agent@phenotype.ai>

- forgecode: deleted 4 stale feat/* local branches (merged or >30 days old) - forgecode: pruned 40 stale remote-tracking refs via git remote prune origin - KooshaPari/phenodag: closed 2 stale PRs (#16, #17) — 8 days old, CI failing or blocked by branch protection Co-authored-by: Phenotype Agent <agent@phenotype.ai>

…only impl (#79) * feat(repo-map): tree-sitter repository structure map Adds forge_repo_map crate — a tree-sitter-based module that walks a repository, parses source files with tree-sitter, and builds a compact symbol/structure map (files -> top-level functions, types, imports) that an agent can use as repo context. Supports Rust, TypeScript, JavaScript, Python, Go with tree-sitter grammars. gitignore-aware file walking via ignore crate. Builder API with max_files cap, configurable cwd, and stable text output format. Adds 19 tests covering: - Language detection (6 langs, fallback) - Rust: function, struct, enum, trait, type alias, impl, use import, unsafe - TypeScript: function, interface, class, type alias, enum, import - Python: function, class, import - Go: function, struct, interface, method, type alias, import - Builder: file discovery, gitignore respect, max_files, stable ordering, unsupported extension filtering, empty directory, text output format * test(dispatch): add ENV_LOCK mutex for parallel test isolation feat(forge3d): add clap + tracing-subscriber workspace deps * chore(sbom): add CycloneDX SBOM artifacts + session docs * feat(forge3d,similarity): forge3d CLI entry + hash-only similarity impl --------- Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Add a feature-gated forge-dev [[bin]] target alongside the existing forge binary so operators can install the Phenotype build side-by-side with the upstream release. The forge-dev target is opt-in via the empty dev-binary feature flag, keeping the default cargo install behavior unchanged. - crates/forge_main/Cargo.toml: declare [features] dev-binary = [] and a new [[bin]] name = "forge-dev" with required-features = ["dev-binary"]. The shared src/main.rs path produces the documented dual-target cargo warning, which is expected. - docs/forge-dev-install.md: 2-paragraph install guide covering the side-by-side rationale, the --features dev-binary gate, and the cargo install command from the git fork. cargo check -p forge_main defaults to building only the forge bin (no warning). cargo check -p forge_main --features dev-binary builds both forge and forge-dev bins and finishes clean. Refs: docs/packaging/FORGE_DEV_PACKAGING.md §2.1, §6.5. Co-authored-by: ForgeCode Bot <forgecode-bot@users.noreply.github.com> Co-authored-by: ForgeCode <noreply@forgecode.dev>

…ommands (T57) (#82) - Add version = env!("CARGO_PKG_VERSION") to #[command(...)]. - Add long_about describing the full daemon startup sequence (pidfile acquisition, drift detector init, JSON-RPC socket bind), the already-running delegate-ping behaviour, and graceful shutdown. - Expand field doc comments: drift_dir notes writable-directory requirement; socket_path explains stale-socket cleanup; pidfile_dir describes flock semantics; forge3_client clarifies both the daemon-running and daemon-not-running paths. Co-authored-by: KooshaPari <koosha@example.com>

* chore(deps): update rust crate config to v0.15.25 (tailcallhq#3574) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.1 (tailcallhq#3575) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency ai to v7.0.3 (tailcallhq#3576) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate posthog-rs to v0.14.2 (tailcallhq#3577) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.2 (tailcallhq#3578) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency ai to v7.0.4 (tailcallhq#3579) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate open to v5.3.6 (tailcallhq#3587) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate indicatif to v0.18.5 (tailcallhq#3589) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * fix(deps): update rust crate posthog-rs to 0.15.0 (tailcallhq#3590) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.3 (tailcallhq#3593) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency ai to v7.0.6 (tailcallhq#3594) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.4 (tailcallhq#3601) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency ai to v7.0.8 (tailcallhq#3597) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate posthog-rs to v0.15.1 (tailcallhq#3602) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore: Claude Sonnet 5 to all providers (tailcallhq#3604) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.5 (tailcallhq#3605) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency ai to v7.0.9 (tailcallhq#3606) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate clap_complete to v4.6.6 (tailcallhq#3607) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * fix: Neuralwatt provider (tailcallhq#3608) * chore(deps): update rust crate console to v0.16.4 (tailcallhq#3609) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.7 (tailcallhq#3612) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency ai to v7.0.11 (tailcallhq#3613) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.8 (tailcallhq#3617) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate clap_complete to v4.6.7 (tailcallhq#3618) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate ignore to v0.4.27 (tailcallhq#3610) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate indicatif to v0.18.6 (tailcallhq#3619) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update rust crate rand to v0.10.2 (tailcallhq#3621) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * fix(deps): update rust crate posthog-rs to 0.17.0 (tailcallhq#3620) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency csv-parse to v7.0.1 (tailcallhq#3625) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency tsx to v4.22.5 (tailcallhq#3626) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency @ai-sdk/google-vertex to v5.0.10 (tailcallhq#3627) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(deps): update dependency ai to v7.0.14 (tailcallhq#3628) Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> * chore(sync): regenerate Cargo.lock after upstream merge (resolves cleanly) --------- Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> Co-authored-by: Amit Singh <amitksingh1490@gmail.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com> Co-authored-by: KooshaPari <koosha@example.com>

Replace the old gh-workflows-generated release.yml (forge binary, manual release trigger) with a cargo-dist-style pipeline that: - Triggers on tag push (v*) - Builds forge-dev binary for mac/linux/windows (5 targets) - Generates install.sh with cross-platform detection - Creates GitHub Release with binaries + installers Adds 'Install forge-dev' section to README with one-line curl install command. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

@2x

) Phenotype-org addition (not present in upstream tailcallhq/forgecode). - New assets/brand/forgecode-icon.svg (1024x1024, hand-authored Terminal-Forge mark) Palette: deep-charcoal #0e0e10, deep-charcoal-2 #1c1c1f, amber-crt #f5a623, synthwave-magenta #d946a8, mint-prompt #6ee7b7. Stylized terminal window with traffic-light dots, amber-CRT 'F>' prompt, magenta spark glow (the AI hint), mint echo line. Subtle CRT scanlines for retro feel. - Rendered iconset: assets/icons/forgecode.iconset/ — 12 PNGs covering Apple's required sizes (16/32/48/64/128/256/512/1024 + @2x variants for 16/32/128/256). Sourced from the SVG via rsvg-convert. - Windows .ico (assets/icons/forgecode.ico) — multi-res 16/32/48/64/128/256 packed via ImageMagick 'convert'. - Linux 256x256.png (assets/icons/forgecode-256x256.png) — standalone. - Wire [package.metadata.bundle] in crates/forge_main/Cargo.toml pointing at the iconset (consumed by tauri/icns bundlers). Marked as Phenotype-org addition per fork CLAUDE.md. - assets/brand/README.md documents palette, family position (distinct from Tracera/MelosViz/Backbone-2/Lab-Coat), and regen snippet. - crates/forge_main/tests/iconset.rs — 11 integration cases covering file presence, palette cohesion, viewBox 1024, Apple sizes + @2x variants, .ico magic bytes, bundle metadata block, AND palette-isolation assertion (no Backbone-2 / Tracera / Lab-Coat hex leaks). 11/11 pass. No AI footer. GPG opt-out. Refs: visual-identity L96 (forgecode 7/36 -> ~17/36) Terminal-Forge palette proposed 2026-07-06 (vision-pillar) fork-conformant per Phenotype Org Cross-Project Reuse Protocol Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Phase 3 visual polish — surface the brand icon at the top of README.md so the canonical visual identity is the first thing renderers see. - substrate: assets/brand/substrate-icon.svg (Backbone-2 golden icon set shipped via PR tailcallhq#221) - forgecode: assets/brand/forgecode-icon.svg (Terminal-Forge golden icon set shipped via PR #87) Additive only — existing prose preserved. No code, no tokens.css edits. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

- Add <repo>-icon-animated.svg (SMIL, no JS, 3.5-5s seamless loop) - Append Motion variant section to brand README - Static icon unchanged - Companion to L96 static golden icon set vision-pillar L101 (Tier-3 polish) Co-authored-by: Phenotype Agent <agent@phenotype.ai>

Terminal-Forge scanline + mint-cursor identity demo. 5s loop rendered via playwright + ffmpeg from CSS-animated SVG. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

… addition) (#86) Upstream fork annotation: this is a Phenotype-org addition. It is NOT present in tailcallhq/forgecode. Adds `crates/forge_display/src/theme.rs` with the Terminal-Forge family palette per the vision-pillar L96 lock-in: tf-deep-charcoal #0d1117 (panel / background) tf-deep-charcoal-2 #161b22 (nested panel / surface) tf-amber-crt #ffb454 (forgecode dominant) tf-synthwave #ff7edb (accent) tf-mint-prompt #7ee787 (success / prompt) Exports: - `TerminalForgePalette` (struct with hex-string fields, Copy) - `terminal_skin_from_theme(&TerminalForgePalette) -> MadSkin` helper that wires the palette into termimad's MadSkin so MarkdownFormat (and downstream consumers) inherit Terminal-Forge identity without per-call-site color literals. Wires `pub mod theme;` into forge_display/src/lib.rs and re-exports the public symbols. Tests (+4): - terminal_forge_palette_matches_tokens_css_family_3: pins the L96 decision (any drift must come from coordinated tokens.css update) - parse_hex_accepts_canonical_terminal_forge_tokens: 5 hex inputs - parse_hex_rejects_malformed_input: 4 negative cases - terminal_skin_from_theme_produces_termimad_skin: smoke-test MadSkin wire-up Local gate: cargo test -p forge_display → 29 tests pass, 0 fail (was 25 before this PR). Co-authored-by: Phenotype Agent <agent@phenotype.ai>

…#91) Co-authored-by: Phenotype Agent <agent@phenotype.ai>

* feat(packaging): forge-dev binary + install guide (T18) Add a feature-gated forge-dev [[bin]] target alongside the existing forge binary so operators can install the Phenotype build side-by-side with the upstream release. The forge-dev target is opt-in via the empty dev-binary feature flag, keeping the default cargo install behavior unchanged. - crates/forge_main/Cargo.toml: declare [features] dev-binary = [] and a new [[bin]] name = "forge-dev" with required-features = ["dev-binary"]. The shared src/main.rs path produces the documented dual-target cargo warning, which is expected. - docs/forge-dev-install.md: 2-paragraph install guide covering the side-by-side rationale, the --features dev-binary gate, and the cargo install command from the git fork. cargo check -p forge_main defaults to building only the forge bin (no warning). cargo check -p forge_main --features dev-binary builds both forge and forge-dev bins and finishes clean. Refs: docs/packaging/FORGE_DEV_PACKAGING.md §2.1, §6.5. Co-Authored-By: ForgeCode <noreply@forgecode.dev> * audit(v38): forgecode C03 scorecard — Agent Readiness L30.1-L30.12 Authored by parent (not codex) because the codex standalone dispatch stalled at the read-evidence phase (log 161KB of source-file reads, no file written, no proc visible after 3 minutes idle). Same pattern as the earlier Tracera failure. DEBUG: every pillar cites real file:line refs (AGENTS.md:1-246, CLAUDE.md:1-133, 2888 #[test] markers, etc.). OPTIMIZE: 12 pillars × ~10 lines each ≈ 120 lines total. ENRICH: every gap has explicit effort (S/M) for follow-up. CLUSTER_TOTAL score=18/36 pct=50% grade=D * audit(v38): forgecode C05 scorecard — 5/30 F * audit(v38): forgecode C06 scorecard — 12/24 C * audit(v38): forgecode C07 scorecard — 16/30 C * audit(v38): forgecode C04 scorecard — Security (18/30) * audit(v38): forgecode C08 scorecard — Eval Coverage (8/30 ) * audit(v38): forgecode C09 scorecard — Accessibility+UX (23/42 ) * audit(v38): forgecode C10 scorecard () * audit(v38): forgecode C09 scorecard (soft_goal_delta: friction capture is not yet instrumented18/42 ) * audit(v38): forgecode C10 scorecard (10/33 ) * audit(v38): forgecode scorecard (audit/.lane-c11/C11.md:7/30 ) Parent-authored (continuation of v38 scorecard sweep). * audit(v38): forgecode C03.md refinement (25/36 ) Re-scan refinement with additional evidence. CLUSTER_TOTAL marker preserved. * audit(v38): forgecode C05.md refinement (5/30 ) Re-scan refinement with additional evidence. CLUSTER_TOTAL marker preserved. --------- Co-authored-by: ForgeCode Bot <forgecode-bot@users.noreply.github.com> Co-authored-by: ForgeCode <noreply@forgecode.dev> Co-authored-by: Phenotype Agent <agent@phenotype.ai>

…se policy (#92) * renames(helioslite): add canonical binary + publish-surface metadata + provenance docs * renames(helioslite): update channel + doctor banner + packaging matrix + landing/Caddy * ci(helioslite): add main branch-protection config (file-only) * renames(helioslite): redirect banner + env-var legacy fallback + tombstone * docs(helioslite): add redirect-chain + install-time tombstone sections

- get_all_provider_models now gracefully handles individual provider failures instead of failing the entire /models endpoint when one provider returns a 500. Individual errors are logged as warnings; the operation only fails if every configured provider fails. - Added /providers remove TUI command (AppCommand::ProviderRemove) that lets users interactively select and remove a configured provider from the REPL, matching the behavior of the existing /logout command.

CLAassistant · 2026-07-15T02:35:30Z

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you all sign our Contributor License Agreement before we can accept your contribution.
0 out of 2 committers have signed the CLA.

❌ Phenotype Agent
❌ KooshaPari

Phenotype Agent seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

KooshaPari and others added 30 commits May 1, 2026 20:06

docs: add CLAUDE.md with fork context and Phenotype-org additions

6e6679c

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

docs: add journey-traceability + iconography implementation (#4)

af78ca8

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

chore: update stale workflow

2fdbc96

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

ci: add trufflehog secrets scan

a13e279

docs: add CODEOWNERS

0ca21da

ci: add FUNDING.yml

e322549

docs: add CONTRIBUTING.md and SECURITY.md

f04b643

Add CHANGELOG.md scaffolding

2658d36

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

ci(cargo-deny): add workflow_dispatch trigger, fix double-tag checkout

208dce5

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

chore: bump version to 0.1.1 (#6)

1d7024f

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

chore: commit untracked infrastructure files

4a97dce

ci: pin all GitHub Actions SHA [org-bootstrap-2026-05-03]

2d0cd6d

fix(cargo-deny): remove stale RUSTSEC-2026-0049 ignore (#7)

864dd7c

Verified resolved upstream; advisory no longer triggers. Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

fix(forgecode): update compact history handling

8fd7740

security(ci): replace trufflehog/actions/setup with go install + setu…

ebd253b

…p-go Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

fix(forgecode): restore workspace members list

c14fe20

Restore the workspace members array by listing all crate directories present on disk, removing invalid glob patterns. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

chore: add cargo-deny bans config and update workflows

da8df12

- Add [bans] section with recommended warnings - Update GitHub workflow files (trufflehog, stale, labels, release)

fix(forgecode): restore workspace members

1e757df

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

fix(workflows): use pull_request instead of pull_request_target

b2b6383

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

chore: add omniroute benchmark plan

bfbfba7

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

chore(forgecode): add packageManager field

9bdfc16

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

fix(rust): tighten deny.toml wildcards policy in forgecode

1b59732

KooshaPari and others added 27 commits June 28, 2026 22:53

chore: upstream diff analysis report (#70)

aefcabc

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

feat(forge-tui): wire forge_tui TUI dashboard crate into workspace (#80)

465542a

Co-authored-by: Phenotype Agent <agent@phenotype.ai>

docs(identity): L105 animated demo media (SVG + MP4 + README) (#90)

ca824ff

Terminal-Forge scanline + mint-cursor identity demo. 5s loop rendered via playwright + ffmpeg from CSS-animated SVG. Co-authored-by: Phenotype Agent <agent@phenotype.ai>

docs(readme): L107 — link to visual identity demo (L105 ladder close) (…

f5fc938

…#91) Co-authored-by: Phenotype Agent <agent@phenotype.ai>

github-actions Bot added type: fix Iterations on existing features or infrastructure. type: provider Updates provider.json configuration. labels Jul 15, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix: per-provider error resilience and /providers remove TUI command#3685

fix: per-provider error resilience and /providers remove TUI command#3685
KooshaPari wants to merge 120 commits into
tailcallhq:mainfrom
KooshaPari:fix/per-provider-error-resilience

KooshaPari commented Jul 15, 2026

Uh oh!

CLAassistant commented Jul 15, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

KooshaPari commented Jul 15, 2026

SummaryTwo fixes addressing provider error handling in the forge TUI:

1. Per-provider error resilience in /models

2. /providers remove TUI command

Changes

Uh oh!

CLAassistant commented Jul 15, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

2. `/providers remove` TUI command