@piagent/platform 1.6.0 → 1.6.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +49 -0
- package/README.md +17 -14
- package/README.vi.md +26 -1
- package/benchmarks/production-v1/spend-control.v1.json +27 -3
- package/benchmarks/production-v1/suite.json +5 -1
- package/catalog/capabilities.json +16 -16
- package/docs/authority-manifest.md +1 -1
- package/docs/auto-delegation-policy.md +38 -73
- package/docs/command-reference-vietnamese.md +23 -31
- package/docs/context-engine.md +5 -3
- package/docs/context-window-policy.md +37 -1
- package/docs/distribution-standard.md +7 -7
- package/docs/en/architecture.md +26 -13
- package/docs/languages.json +4 -4
- package/docs/operator-manual-vietnamese.md +21 -41
- package/docs/publishing-for-teams.md +5 -0
- package/docs/quality-benchmark.md +179 -23
- package/docs/release-install-policy.md +6 -6
- package/docs/runtime-harness-standard.md +3 -3
- package/docs/security-threat-model.md +1 -1
- package/docs/subagent-orchestration-capabilities.md +19 -51
- package/docs/subagents-and-multiagent.md +21 -47
- package/docs/task-implementation-contract.md +3 -3
- package/docs/task-lifecycle-vietnamese.md +1 -1
- package/docs/team-onboarding.md +2 -2
- package/docs/vi/architecture.md +26 -13
- package/docs-site/architecture.html +1 -1
- package/docs-site/benchmark.html +89 -95
- package/docs-site/commands.html +1 -1
- package/docs-site/content/benchmark.html +88 -94
- package/docs-site/content/en/benchmark.html +134 -90
- package/docs-site/content/en/index.html +19 -19
- package/docs-site/content/en/runtime.html +4 -4
- package/docs-site/content/en/whats-new.html +12 -1
- package/docs-site/content/en/workflows.html +2 -2
- package/docs-site/content/index.html +19 -19
- package/docs-site/content/runtime.html +4 -4
- package/docs-site/content/whats-new.html +12 -1
- package/docs-site/content/workflows.html +2 -2
- package/docs-site/context-engine.html +1 -1
- package/docs-site/en/architecture.html +1 -1
- package/docs-site/en/benchmark.html +135 -91
- package/docs-site/en/commands.html +1 -1
- package/docs-site/en/context-engine.html +1 -1
- package/docs-site/en/index.html +20 -20
- package/docs-site/en/mcp-approval.html +1 -1
- package/docs-site/en/mcp-auth.html +1 -1
- package/docs-site/en/mcp-servers.html +1 -1
- package/docs-site/en/mcp.html +1 -1
- package/docs-site/en/permissions.html +1 -1
- package/docs-site/en/profiles.html +1 -1
- package/docs-site/en/quickstart.html +1 -1
- package/docs-site/en/runtime.html +5 -5
- package/docs-site/en/scope.html +1 -1
- package/docs-site/en/security.html +1 -1
- package/docs-site/en/team.html +1 -1
- package/docs-site/en/whats-new.html +13 -2
- package/docs-site/en/workflows.html +3 -3
- package/docs-site/index.html +20 -20
- package/docs-site/mcp-approval.html +1 -1
- package/docs-site/mcp-auth.html +1 -1
- package/docs-site/mcp-servers.html +1 -1
- package/docs-site/mcp.html +1 -1
- package/docs-site/permissions.html +1 -1
- package/docs-site/profiles.html +1 -1
- package/docs-site/quickstart.html +1 -1
- package/docs-site/runtime.html +5 -5
- package/docs-site/scope.html +1 -1
- package/docs-site/security.html +1 -1
- package/docs-site/team.html +1 -1
- package/docs-site/whats-new.html +13 -2
- package/docs-site/workflows.html +3 -3
- package/evals/fixtures/benchmark-assurance-evidence.invalid.json +1 -1
- package/evals/fixtures/benchmark-assurance-evidence.valid.json +1 -1
- package/evals/fixtures/codex-relative-efficiency-v1.invalid.json +4 -0
- package/evals/fixtures/codex-relative-efficiency-v1.valid.json +94 -0
- package/evals/fixtures/helper-request.invalid.json +1 -1
- package/evals/fixtures/helper-request.valid.json +1 -1
- package/evals/fixtures/piagent-webui/subagent-tree-v1.valid.json +7 -0
- package/evals/fixtures/task-authority-snapshot.valid.json +4 -4
- package/evals/fs4-readiness-matrix.v1.json +2 -2
- package/evals/fs5-causal-arm.v1.json +2 -2
- package/evals/fs5-pilot-protocol.v1.json +3 -3
- package/evals/fs5-pilot-protocol.v2.json +8 -8
- package/evals/fs5-pilot-protocol.v3.json +8 -8
- package/evals/fs5-pilot-protocol.v4.json +6 -6
- package/evals/fs5-pilot-protocol.v5.json +7 -6
- package/evals/private-holdout-v1/public-exposure.v1.json +1 -1
- package/evals/runtime-conformance-v1/README.md +43 -0
- package/evals/runtime-conformance-v1/evaluate-paid-canary.mjs +198 -0
- package/evals/runtime-conformance-v1/lane.json +57 -0
- package/evals/runtime-conformance-v1/paid-canary.json +31 -0
- package/evals/runtime-conformance-v1/runner.mjs +456 -0
- package/package.json +1 -1
- package/packages/piagent-core/README.md +13 -4
- package/packages/piagent-core/benchmark/benchmark-codex-relative-efficiency.js +698 -0
- package/packages/piagent-core/benchmark/benchmark-core.js +4 -3
- package/packages/piagent-core/benchmark/benchmark-forensics.js +4 -2
- package/packages/piagent-core/benchmark/benchmark-preflight.js +3 -1
- package/packages/piagent-core/benchmark/benchmark-provider-free-evidence.js +345 -0
- package/packages/piagent-core/benchmark/benchmark-record-validation.js +74 -5
- package/packages/piagent-core/benchmark/benchmark-report.js +15 -1
- package/packages/piagent-core/benchmark/benchmark-runtime-causal.js +244 -0
- package/packages/piagent-core/benchmark/benchmark-stage-diagnostic.js +201 -4
- package/packages/piagent-core/benchmark/benchmark-suite.js +22 -1
- package/packages/piagent-core/benchmark/benchmark-summary-support.js +4 -0
- package/packages/piagent-core/benchmark/benchmark-usage.js +15 -4
- package/packages/piagent-core/capabilities/runtime-integrity.js +3 -0
- package/packages/piagent-core/extensions/acceptance-behavior-proof.js +806 -0
- package/packages/piagent-core/extensions/acceptance-receipt.js +89 -48
- package/packages/piagent-core/extensions/context-engine.js +1 -11
- package/packages/piagent-core/extensions/context-index-policy.js +4 -0
- package/packages/piagent-core/extensions/context-query-paths.js +57 -0
- package/packages/piagent-core/extensions/criterion-context-pack.js +16 -3
- package/packages/piagent-core/extensions/criterion-graph.js +4 -1
- package/packages/piagent-core/extensions/fallback-base-policy.ts +101 -0
- package/packages/piagent-core/extensions/guard-types.ts +4 -0
- package/packages/piagent-core/extensions/piagent-guard.ts +194 -226
- package/packages/piagent-core/extensions/task-lifecycle.js +32 -0
- package/packages/piagent-core/extensions/task-state.js +69 -8
- package/packages/piagent-core/extensions/verification-intelligence.js +49 -12
- package/packages/piagent-core/extensions/workspace-evidence-roots.js +52 -16
- package/packages/piagent-core/package.json +1 -1
- package/packages/piagent-core/policies/base-policy.json +42 -2
- package/packages/piagent-core/policy/authority-manifest.v1.json +3 -3
- package/packages/piagent-core/prompts/be-to-fe.md +2 -8
- package/packages/piagent-core/prompts/plan.md +1 -1
- package/packages/piagent-core/prompts/platform-improve.md +2 -6
- package/packages/piagent-core/prompts/review.md +2 -7
- package/packages/piagent-core/prompts/scout.md +1 -1
- package/packages/piagent-core/prompts/task.md +4 -3
- package/packages/piagent-core/runtime/hooks/completion-hook.ts +7 -6
- package/packages/piagent-core/runtime/hooks/input-hook.ts +6 -0
- package/packages/piagent-core/runtime/hooks/session-start-hook.ts +100 -1
- package/packages/piagent-core/runtime/hooks/tool-result-hook.ts +58 -72
- package/packages/piagent-core/runtime/hooks/tool-result-value-helpers.ts +66 -0
- package/packages/piagent-core/runtime/inspection/bounded-emission-guard.ts +286 -0
- package/packages/piagent-core/runtime/inspection/runtime-event-store.ts +37 -7
- package/packages/piagent-core/runtime/inspection/task-subagent-tree.ts +27 -4
- package/packages/piagent-core/runtime/inspection/tool-failure-classification.ts +20 -3
- package/packages/piagent-core/runtime/inspection/webui-snapshot.ts +47 -11
- package/packages/piagent-core/runtime/orchestration/helper-lifecycle.ts +16 -12
- package/packages/piagent-core/runtime/orchestration/minimal-delegation-policy.ts +71 -0
- package/packages/piagent-core/runtime/orchestration/owned-work-budget.ts +28 -20
- package/packages/piagent-core/runtime/orchestration/role-policy.ts +27 -9
- package/packages/piagent-core/runtime/orchestration/subagent-tool-policy.ts +329 -0
- package/packages/piagent-core/runtime/policy/block-remedy.ts +0 -4
- package/packages/piagent-core/runtime/product/efficiency-metrics.ts +3 -0
- package/packages/piagent-core/runtime/quality/edit-freshness-guard.ts +102 -0
- package/packages/piagent-core/runtime/quality/exact-output-contract.ts +20 -10
- package/packages/piagent-core/runtime/quality/performance-assurance.ts +3 -3
- package/packages/piagent-core/runtime/recovery/continuation-budget.ts +1 -1
- package/packages/piagent-core/runtime/recovery/handoff-projection.ts +4 -1
- package/packages/piagent-core/runtime/recovery/pristine-task-policy-refresh.ts +91 -0
- package/packages/piagent-core/runtime/recovery/recovery-policy.ts +7 -4
- package/packages/piagent-core/runtime/recovery/resume-state.ts +112 -14
- package/packages/piagent-core/runtime/recovery/semantic-repair-handshake.ts +7 -13
- package/packages/piagent-core/runtime/recovery/task-evidence-coverage-refresh.ts +217 -0
- package/packages/piagent-core/runtime/registration/context-commands.ts +16 -2
- package/packages/piagent-core/runtime/registration/knowledge-tools.ts +9 -5
- package/packages/piagent-core/runtime/registration/operator-request-intake.ts +49 -0
- package/packages/piagent-core/runtime/registration/policy-tools.ts +4 -4
- package/packages/piagent-core/runtime/registration/session-commands.ts +1 -1
- package/packages/piagent-core/runtime/registration/task-completion-tools.ts +12 -9
- package/packages/piagent-core/runtime/registration/task-evidence-tools.ts +21 -12
- package/packages/piagent-core/runtime/registration/task-start-guidance.ts +19 -13
- package/packages/piagent-core/runtime/registration/task-start-manifest.ts +54 -0
- package/packages/piagent-core/runtime/registration/task-start-tool.ts +57 -57
- package/packages/piagent-core/runtime/runtime-limits.ts +16 -2
- package/packages/piagent-core/runtime/session/adaptive-context-analysis.ts +461 -0
- package/packages/piagent-core/runtime/session/adaptive-context-governor.ts +464 -0
- package/packages/piagent-core/runtime/session/adaptive-context-ledger.ts +265 -0
- package/packages/piagent-core/runtime/session/message-signals.ts +7 -3
- package/packages/piagent-core/runtime/session/operator-request-carry.ts +44 -0
- package/packages/piagent-core/runtime/session/runtime-state.ts +11 -7
- package/packages/piagent-core/runtime/session/source-checkout-read-grants.ts +35 -0
- package/packages/piagent-core/runtime/session/system-prompt.ts +76 -14
- package/packages/piagent-core/runtime/session/task-contract-presentation.ts +39 -0
- package/packages/piagent-core/runtime/tools/phase-tools.ts +8 -8
- package/packages/piagent-core/runtime/tools/tool-groups.ts +12 -2
- package/packages/piagent-core/runtime/verification/exact-verifier-reuse.ts +101 -0
- package/packages/piagent-core/runtime/workflows/runtime-intake-compaction.ts +164 -0
- package/packages/piagent-core/runtime/workflows/task-intake.ts +167 -39
- package/packages/piagent-core/subagents/piagent-oracle.md +5 -3
- package/packages/piagent-core/subagents/piagent-planner.md +6 -4
- package/packages/piagent-core/subagents/piagent-reviewer.md +5 -3
- package/packages/piagent-core/subagents/piagent-scout.md +4 -2
- package/packages/piagent-core/subagents/piagent-worker.md +6 -4
- package/packages/piagent-webui/dist/client/assets/{MarkdownMessage-CxeZqTte.js → MarkdownMessage-B4XehTCU.js} +1 -1
- package/packages/piagent-webui/dist/client/assets/{index-DynIMe8I.js → index-CgOEyRPG.js} +4 -4
- package/packages/piagent-webui/dist/client/index.html +1 -1
- package/packages/piagent-webui/extension/piagent-webui.ts +3 -1
- package/packages/piagent-webui/extension/same-session-bridge.ts +1 -1
- package/packages/piagent-webui/extension/session-stream-adapter.ts +33 -10
- package/packages/piagent-webui/gateway/gateway-session-stream.ts +37 -21
- package/packages/piagent-webui/gateway/session-operation-lifecycle.ts +237 -0
- package/packages/piagent-webui/gateway/session-operation-retry-control.ts +38 -0
- package/packages/piagent-webui/gateway/session-operation-watchdog.ts +9 -2
- package/packages/piagent-webui/gateway/session-runtime-supervisor.ts +8 -7
- package/packages/piagent-webui/ownership/session-lease-store.ts +5 -2
- package/packages/piagent-webui/server/core-inspection-provider.ts +2 -1
- package/packages/piagent-webui/server/transcript-projection.ts +91 -5
- package/schemas/codex-relative-efficiency-v1.schema.json +155 -0
- package/schemas/piagent-webui/runtime-event-v2.schema.json +15 -5
- package/schemas/piagent-webui/subagent-tree-v1.schema.json +11 -1
- package/schemas/project-profile.schema.json +1 -1
- package/schemas/task-contract.schema.json +15 -0
- package/scripts/benchmark-runner-core.mjs +14 -6
- package/scripts/benchmark-runner-finalization.mjs +4 -0
- package/scripts/configure-subagents.sh +42 -34
- package/scripts/helper-evaluation.mjs +1 -1
- package/scripts/team-doctor.sh +1 -1
- package/scripts/verify-local.sh +5 -0
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · EN</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Language">
|
|
35
35
|
<a class="language-option" href="/quickstart" hreflang="vi" lang="vi">VI</a>
|
|
36
36
|
<a class="language-option current" href="/en/quickstart" hreflang="en" lang="en" aria-current="true">EN</a>
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · EN</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Language">
|
|
35
35
|
<a class="language-option" href="/runtime" hreflang="vi" lang="vi">VI</a>
|
|
36
36
|
<a class="language-option current" href="/en/runtime" hreflang="en" lang="en" aria-current="true">EN</a>
|
|
@@ -115,11 +115,11 @@ piagent-install --version vX.Y.Z --resolve-tag</code></pre></article>
|
|
|
115
115
|
<article class="command-card"><header><span>Doctor</span><button class="clip-btn" data-clip="piagent-doctor /path/to/project">Copy</button></header><pre><code>piagent-doctor /path/to/project</code></pre></article>
|
|
116
116
|
<article class="command-card"><header><span>Smoke benchmark</span><button class="clip-btn" data-clip="piagent-benchmark --dry-run piagent-benchmark">Copy</button></header><pre><code>piagent-benchmark --dry-run
|
|
117
117
|
piagent-benchmark</code></pre></article>
|
|
118
|
-
<article class="command-card"><header><span>Production benchmark</span><button class="clip-btn" data-clip="piagent-benchmark --production --dry-run piagent-benchmark --production --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-
|
|
118
|
+
<article class="command-card"><header><span>Production benchmark</span><button class="clip-btn" data-clip="piagent-benchmark --production --dry-run piagent-benchmark --production --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-luna --thinking medium">Copy</button></header><pre><code>piagent-benchmark --production --dry-run
|
|
119
119
|
piagent-benchmark --production --surfaces piagent,codex-cli \
|
|
120
|
-
--model openai-codex/gpt-5.6-
|
|
121
|
-
<article class="command-card"><header><span>Compare codex-cli</span><button class="clip-btn" data-clip="piagent-benchmark --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-
|
|
122
|
-
--model openai-codex/gpt-5.6-
|
|
120
|
+
--model openai-codex/gpt-5.6-luna --thinking medium</code></pre></article>
|
|
121
|
+
<article class="command-card"><header><span>Compare codex-cli</span><button class="clip-btn" data-clip="piagent-benchmark --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-luna --thinking medium">Copy</button></header><pre><code>piagent-benchmark --surfaces piagent,codex-cli \
|
|
122
|
+
--model openai-codex/gpt-5.6-luna --thinking medium</code></pre></article>
|
|
123
123
|
<article class="command-card"><header><span>Trusted run</span><button class="clip-btn" data-clip="piagent-auto --full-access -p "Run the trusted local benchmark suite."">Copy</button></header><pre><code>piagent-auto --read-only -p "Scout payment mapping. Do not edit source."
|
|
124
124
|
piagent-auto --full-access -p "Run the trusted local benchmark suite."</code></pre></article>
|
|
125
125
|
<article class="command-card"><header><span>Capabilities</span><button class="clip-btn" data-clip="piagent-capabilities catalog --check piagent-capabilities doctor --profile .pi/piagent-profile.json">Copy</button></header><pre><code>piagent-capabilities catalog --check
|
package/docs-site/en/scope.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · EN</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Language">
|
|
35
35
|
<a class="language-option" href="/scope" hreflang="vi" lang="vi">VI</a>
|
|
36
36
|
<a class="language-option current" href="/en/scope" hreflang="en" lang="en" aria-current="true">EN</a>
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · EN</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Language">
|
|
35
35
|
<a class="language-option" href="/security" hreflang="vi" lang="vi">VI</a>
|
|
36
36
|
<a class="language-option current" href="/en/security" hreflang="en" lang="en" aria-current="true">EN</a>
|
package/docs-site/en/team.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · EN</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Language">
|
|
35
35
|
<a class="language-option" href="/team" hreflang="vi" lang="vi">VI</a>
|
|
36
36
|
<a class="language-option current" href="/en/team" hreflang="en" lang="en" aria-current="true">EN</a>
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · EN</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Language">
|
|
35
35
|
<a class="language-option" href="/whats-new" hreflang="vi" lang="vi">VI</a>
|
|
36
36
|
<a class="language-option current" href="/en/whats-new" hreflang="en" lang="en" aria-current="true">EN</a>
|
|
@@ -102,6 +102,17 @@
|
|
|
102
102
|
<h1>Safer task flow with less manual coordination</h1>
|
|
103
103
|
<p>Product stabilization adds runtime truth, trajectories, recovery, helper policy, an activity inspector, and fail-closed completion receipts without breaking existing workflows.</p>
|
|
104
104
|
</div>
|
|
105
|
+
<section class="section">
|
|
106
|
+
<p class="section-kicker">v1.6.1 · Reliability patch</p>
|
|
107
|
+
<h2 id="v1-6-1">Truthful UI state, safe external review, and lower helper overhead</h2>
|
|
108
|
+
<p class="section-lead">This patch keeps the selected model and thinking level unchanged while tightening lifecycle truth, limiting helper expansion, and granting external-source reads only to the exact checkout for the current session.</p>
|
|
109
|
+
<div class="grid cols-3">
|
|
110
|
+
<article class="card"><h3 id="conversation-truth">Conversation truth</h3><p>Drafts, retries, and pending completion stay in Activity. Chat preserves user-then-assistant order and clears processing only at the real terminal boundary.</p></article>
|
|
111
|
+
<article class="card"><h3 id="external-review">External review</h3><p>GitHub, GitLab, and Bitbucket references activate only minimal source checkout. The runtime grants reads to that exact tree while blocking caches, credentials, symlink escape, and every mutation path.</p></article>
|
|
112
|
+
<article class="card"><h3 id="solo-first">Solo-first</h3><p>Direct subagent use is capped at one read-only helper with call, turn, and runtime budgets, no automatic retry/fork/worker expansion, and truthful failure state when evidence is missing.</p></article>
|
|
113
|
+
</div>
|
|
114
|
+
</section>
|
|
115
|
+
|
|
105
116
|
<section class="section">
|
|
106
117
|
<p class="section-kicker">Product stabilization</p>
|
|
107
118
|
<h2 id="main-changes">Six changes users will notice</h2>
|
|
@@ -175,7 +186,7 @@
|
|
|
175
186
|
<tr><td>Solver</td><td><span class="inline-code">shadow</span></td><td>Evaluates routes without changing the parent model or taking control of the workflow.</td></tr>
|
|
176
187
|
<tr><td>Phase tools</td><td><span class="inline-code">shadow</span></td><td>Computes projections; use <span class="inline-code">PIAGENT_PHASE_TOOLS=on</span> to apply phase-specific tool visibility.</td></tr>
|
|
177
188
|
<tr><td>Recovery</td><td><span class="inline-code">on</span></td><td>Retries or repairs within ceilings; use <span class="inline-code">PIAGENT_AUTO_RECOVERY=off</span> to restore ordinary handoff.</td></tr>
|
|
178
|
-
<tr><td>Helpers</td><td><span class="inline-code">
|
|
189
|
+
<tr><td>Helpers</td><td><span class="inline-code">off</span></td><td>Explicit <span class="inline-code">PIAGENT_HELPERS_MODE=on</span> still allows only one fresh read-only helper after two independent lanes and at least 30% projected net saving are proven.</td></tr>
|
|
179
190
|
<tr><td>Parent routing and automatic workers</td><td><span class="inline-code">off</span></td><td>No automatic parent-model switch, writer dispatch, or default multi-writer behavior.</td></tr>
|
|
180
191
|
</tbody>
|
|
181
192
|
</table>
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · EN</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Language">
|
|
35
35
|
<a class="language-option" href="/workflows" hreflang="vi" lang="vi">VI</a>
|
|
36
36
|
<a class="language-option current" href="/en/workflows" hreflang="en" lang="en" aria-current="true">EN</a>
|
|
@@ -168,8 +168,8 @@
|
|
|
168
168
|
<p>There is no separate <span class="inline-code">/git</span> namespace. Use <span class="inline-code">/workflow commit</span> for a reviewed local commit and <span class="inline-code">/workflow pr</span> to prepare a pull request. Actions such as <span class="inline-code">git push</span> or broad staging with <span class="inline-code">git add -A</span> still require operator confirmation.</p>
|
|
169
169
|
</div>
|
|
170
170
|
<div class="callout">
|
|
171
|
-
<strong>
|
|
172
|
-
<p>Use <span class="inline-code">/piagent-orchestration</span> to inspect
|
|
171
|
+
<strong>Parent-direct orchestration</strong>
|
|
172
|
+
<p>Use <span class="inline-code">/piagent-orchestration</span> to inspect the one-helper ceiling, dispatch/skip reason, projected token saving, review lenses, Field Guide, and writer policy. Helpers default to off. Explicit opt-in still permits only one fresh read-only helper when two independent lanes and at least 30% projected net saving are proven; workers and retries stay disabled.</p>
|
|
173
173
|
</div>
|
|
174
174
|
<div class="command-grid">
|
|
175
175
|
<article class="command-card"><header><span>Bounded implementation</span><button class="clip-btn" data-clip="/workflow task Fix the failing checkout test and verify.">Copy</button></header><pre><code>/workflow task Fix the failing checkout test and verify.</code></pre></article>
|
package/docs-site/index.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/" hreflang="en" lang="en">EN</a>
|
|
@@ -112,30 +112,30 @@
|
|
|
112
112
|
</section>
|
|
113
113
|
|
|
114
114
|
<section class="section">
|
|
115
|
-
<p class="section-kicker">Production V1 ·
|
|
115
|
+
<p class="section-kicker">Production V1 · bằng chứng public của v1.6.0</p>
|
|
116
116
|
<h2 id="bang-chung">Đo được, không phải nói suông</h2>
|
|
117
|
-
<p class="section-lead">
|
|
117
|
+
<p class="section-lead">Run gắn với release <code>production-v1-20260824T040017Z-05b7cf</code> đã đo 108 session bằng GPT-5.6 Luna Medium tại tag <code>v1.6.0</code>, đúng commit <code>3bba8f0b3ff521bc2a355e1f6bef6d1bbdc09511</code>, với Piagent so với <code>codex-cli</code>.</p>
|
|
118
118
|
|
|
119
119
|
<div class="benchmark-metrics" aria-label="Số liệu benchmark nổi bật">
|
|
120
120
|
<div class="benchmark-metric">
|
|
121
|
-
<span>
|
|
122
|
-
<strong>
|
|
123
|
-
<small
|
|
121
|
+
<span>Bằng chứng release</span>
|
|
122
|
+
<strong>v1.6.0</strong>
|
|
123
|
+
<small>đúng commit 3bba8f0</small>
|
|
124
124
|
</div>
|
|
125
125
|
<div class="benchmark-metric">
|
|
126
126
|
<span>Task resolved</span>
|
|
127
127
|
<strong>54/54</strong>
|
|
128
|
-
<small>
|
|
128
|
+
<small>Piagent; codex-cli đạt 48/54</small>
|
|
129
129
|
</div>
|
|
130
130
|
<div class="benchmark-metric">
|
|
131
131
|
<span>Fresh token</span>
|
|
132
|
-
<strong>-
|
|
133
|
-
<small>
|
|
132
|
+
<strong>-61,43%</strong>
|
|
133
|
+
<small>tỷ lệ family fixed-workload chính 0,3857</small>
|
|
134
134
|
</div>
|
|
135
135
|
<div class="benchmark-metric">
|
|
136
|
-
<span>
|
|
137
|
-
<strong>
|
|
138
|
-
<small>
|
|
136
|
+
<span>Bằng chứng usage</span>
|
|
137
|
+
<strong>108/108 exact</strong>
|
|
138
|
+
<small>không có retry</small>
|
|
139
139
|
</div>
|
|
140
140
|
</div>
|
|
141
141
|
|
|
@@ -144,20 +144,20 @@
|
|
|
144
144
|
<span><i class="codex-swatch"></i>codex-cli</span>
|
|
145
145
|
</div>
|
|
146
146
|
|
|
147
|
-
<div class="benchmark-usage-chart" aria-label="
|
|
147
|
+
<div class="benchmark-usage-chart" aria-label="Tỷ lệ fresh token theo family fixed-workload: Piagent 38,57 phần trăm, baseline codex-cli 100 phần trăm; task hoàn thành: Piagent 54 trên 54, codex-cli 48 trên 54">
|
|
148
148
|
<div class="benchmark-usage-group">
|
|
149
|
-
<h3>
|
|
150
|
-
<div class="benchmark-usage-row"><span>Piagent</span><div class="benchmark-track"><i class="piagent-bar" style="--bar-size:
|
|
151
|
-
<div class="benchmark-usage-row"><span>codex-cli</span><div class="benchmark-track"><i class="codex-bar" style="--bar-size: 100%"></i></div><strong>
|
|
149
|
+
<h3>Tỷ lệ fresh token theo family fixed-workload</h3>
|
|
150
|
+
<div class="benchmark-usage-row"><span>Piagent</span><div class="benchmark-track"><i class="piagent-bar" style="--bar-size: 38.57%"></i></div><strong>0,3857</strong></div>
|
|
151
|
+
<div class="benchmark-usage-row"><span>codex-cli</span><div class="benchmark-track"><i class="codex-bar" style="--bar-size: 100%"></i></div><strong>1,0000</strong></div>
|
|
152
152
|
</div>
|
|
153
153
|
<div class="benchmark-usage-group">
|
|
154
|
-
<h3>
|
|
155
|
-
<div class="benchmark-usage-row"><span>Piagent</span><div class="benchmark-track"><i class="piagent-bar" style="--bar-size:
|
|
156
|
-
<div class="benchmark-usage-row"><span>codex-cli</span><div class="benchmark-track"><i class="codex-bar" style="--bar-size:
|
|
154
|
+
<h3>Task resolved</h3>
|
|
155
|
+
<div class="benchmark-usage-row"><span>Piagent</span><div class="benchmark-track"><i class="piagent-bar" style="--bar-size: 100%"></i></div><strong>54/54</strong></div>
|
|
156
|
+
<div class="benchmark-usage-row"><span>codex-cli</span><div class="benchmark-track"><i class="codex-bar" style="--bar-size: 88.89%"></i></div><strong>48/54</strong></div>
|
|
157
157
|
</div>
|
|
158
158
|
</div>
|
|
159
159
|
|
|
160
|
-
<p>
|
|
160
|
+
<p>Tỷ lệ family fixed-workload chính là <strong>0,3857</strong>, với khoảng tin cậy 95% <strong>0,3073–0,4840</strong>. Ước lượng điểm thấp hơn 61,43%; theo cận trên của khoảng tin cậy, có thể phát biểu thận trọng rằng Piagent dùng ít hơn ít nhất 51,60% fresh token trên đúng workload production-v1 đã khai báo trước và đúng release này. Usage của cả 108 session đều exact và không có retry. <a href="/benchmark">Xem phương pháp benchmark và ranh giới bằng chứng</a>.</p>
|
|
161
161
|
</section>
|
|
162
162
|
|
|
163
163
|
<section class="section">
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/mcp-approval" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/mcp-approval" hreflang="en" lang="en">EN</a>
|
package/docs-site/mcp-auth.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/mcp-auth" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/mcp-auth" hreflang="en" lang="en">EN</a>
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/mcp-servers" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/mcp-servers" hreflang="en" lang="en">EN</a>
|
package/docs-site/mcp.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/mcp" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/mcp" hreflang="en" lang="en">EN</a>
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/permissions" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/permissions" hreflang="en" lang="en">EN</a>
|
package/docs-site/profiles.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/profiles" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/profiles" hreflang="en" lang="en">EN</a>
|
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/quickstart" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/quickstart" hreflang="en" lang="en">EN</a>
|
package/docs-site/runtime.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/runtime" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/runtime" hreflang="en" lang="en">EN</a>
|
|
@@ -135,15 +135,15 @@ piagent-install --version vX.Y.Z --resolve-tag</code></pre>
|
|
|
135
135
|
piagent-benchmark</code></pre>
|
|
136
136
|
</article>
|
|
137
137
|
<article class="command-card">
|
|
138
|
-
<header><span>Production benchmark</span><button class="clip-btn" data-clip="piagent-benchmark --production --dry-run piagent-benchmark --production --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-
|
|
138
|
+
<header><span>Production benchmark</span><button class="clip-btn" data-clip="piagent-benchmark --production --dry-run piagent-benchmark --production --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-luna --thinking medium">Lấy lệnh</button></header>
|
|
139
139
|
<pre><code>piagent-benchmark --production --dry-run
|
|
140
140
|
piagent-benchmark --production --surfaces piagent,codex-cli \
|
|
141
|
-
--model openai-codex/gpt-5.6-
|
|
141
|
+
--model openai-codex/gpt-5.6-luna --thinking medium</code></pre>
|
|
142
142
|
</article>
|
|
143
143
|
<article class="command-card">
|
|
144
|
-
<header><span>Compare codex-cli</span><button class="clip-btn" data-clip="piagent-benchmark --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-
|
|
144
|
+
<header><span>Compare codex-cli</span><button class="clip-btn" data-clip="piagent-benchmark --surfaces piagent,codex-cli --model openai-codex/gpt-5.6-luna --thinking medium">Lấy lệnh</button></header>
|
|
145
145
|
<pre><code>piagent-benchmark --surfaces piagent,codex-cli \
|
|
146
|
-
--model openai-codex/gpt-5.6-
|
|
146
|
+
--model openai-codex/gpt-5.6-luna --thinking medium</code></pre>
|
|
147
147
|
</article>
|
|
148
148
|
<article class="command-card">
|
|
149
149
|
<header><span>Trusted run</span><button class="clip-btn" data-clip="piagent-auto --full-access -p "Run the trusted local benchmark suite."">Lấy lệnh</button></header>
|
package/docs-site/scope.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/scope" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/scope" hreflang="en" lang="en">EN</a>
|
package/docs-site/security.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/security" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/security" hreflang="en" lang="en">EN</a>
|
package/docs-site/team.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/team" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/team" hreflang="en" lang="en">EN</a>
|
package/docs-site/whats-new.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/whats-new" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/whats-new" hreflang="en" lang="en">EN</a>
|
|
@@ -102,6 +102,17 @@
|
|
|
102
102
|
<h1>Luồng task an toàn hơn, ít thao tác tay hơn</h1>
|
|
103
103
|
<p>Product stabilization bổ sung runtime truth, trajectory, recovery, helper policy, activity inspector và completion receipt fail-closed mà không phá workflow cũ.</p>
|
|
104
104
|
</div>
|
|
105
|
+
<section class="section">
|
|
106
|
+
<p class="section-kicker">v1.6.1 · Reliability patch</p>
|
|
107
|
+
<h2 id="v1-6-1">UI đúng trạng thái, review repo ngoài an toàn, helper ít tốn hơn</h2>
|
|
108
|
+
<p class="section-lead">Bản vá này giữ nguyên model và thinking level, nhưng siết chặt lifecycle để response chỉ hiện khi task thật sự hoàn tất, đồng thời giới hạn helper và quyền đọc source ngoài theo đúng session.</p>
|
|
109
|
+
<div class="grid cols-3">
|
|
110
|
+
<article class="card"><h3 id="conversation-truth">Conversation truth</h3><p>Draft, retry và completion còn pending ở trong Activity. Chat giữ đúng thứ tự user rồi assistant và không tiếp tục hiện processing sau terminal.</p></article>
|
|
111
|
+
<article class="card"><h3 id="external-review">External review</h3><p>URL GitHub, GitLab và Bitbucket chỉ bật source checkout tối thiểu; runtime cấp quyền đọc chính xác checkout đó và vẫn chặn cache, credential, symlink escape cùng mọi mutation.</p></article>
|
|
112
|
+
<article class="card"><h3 id="solo-first">Solo-first</h3><p>Direct subagent có trần một helper read-only, không auto retry/fork/worker và có giới hạn call, turn, thời gian. Kết quả thiếu bằng chứng được ghi là failure thật.</p></article>
|
|
113
|
+
</div>
|
|
114
|
+
</section>
|
|
115
|
+
|
|
105
116
|
<section class="section">
|
|
106
117
|
<p class="section-kicker">Product stabilization</p>
|
|
107
118
|
<h2 id="thay-doi-chinh">Sáu thay đổi người dùng sẽ thấy</h2>
|
|
@@ -175,7 +186,7 @@
|
|
|
175
186
|
<tr><td>Solver</td><td><span class="inline-code">shadow</span></td><td>Đánh giá route nhưng không đổi parent model hoặc tự chiếm quyền workflow.</td></tr>
|
|
176
187
|
<tr><td>Phase tools</td><td><span class="inline-code">shadow</span></td><td>Tính projection; dùng <span class="inline-code">PIAGENT_PHASE_TOOLS=on</span> để áp tool visibility theo phase.</td></tr>
|
|
177
188
|
<tr><td>Recovery</td><td><span class="inline-code">on</span></td><td>Tự retry/repair trong ceiling; dùng <span class="inline-code">PIAGENT_AUTO_RECOVERY=off</span> để quay về ordinary handoff.</td></tr>
|
|
178
|
-
<tr><td>Helpers</td><td><span class="inline-code">
|
|
189
|
+
<tr><td>Helpers</td><td><span class="inline-code">off</span></td><td><span class="inline-code">PIAGENT_HELPERS_MODE=on</span> vẫn chỉ cho một helper read-only context fresh sau khi chứng minh hai lane độc lập và projected net saving từ 30%.</td></tr>
|
|
179
190
|
<tr><td>Parent routing + automatic worker</td><td><span class="inline-code">off</span></td><td>Không tự đổi parent model, không tự bật writer và không có multi-writer mặc định.</td></tr>
|
|
180
191
|
</tbody>
|
|
181
192
|
</table>
|
package/docs-site/workflows.html
CHANGED
|
@@ -30,7 +30,7 @@
|
|
|
30
30
|
<span>Pi Agent Platform</span>
|
|
31
31
|
</a>
|
|
32
32
|
<div class="top-actions">
|
|
33
|
-
<span class="pill"><span class="pill-dot"></span>v1.6.
|
|
33
|
+
<span class="pill"><span class="pill-dot"></span>v1.6.1 docs · VI</span>
|
|
34
34
|
<nav class="language-switch" aria-label="Ngôn ngữ">
|
|
35
35
|
<a class="language-option current" href="/workflows" hreflang="vi" lang="vi" aria-current="true">VI</a>
|
|
36
36
|
<a class="language-option" href="/en/workflows" hreflang="en" lang="en">EN</a>
|
|
@@ -169,8 +169,8 @@
|
|
|
169
169
|
</div>
|
|
170
170
|
|
|
171
171
|
<div class="callout">
|
|
172
|
-
<strong>
|
|
173
|
-
<p>Dùng <span class="inline-code">/piagent-orchestration</span> để xem
|
|
172
|
+
<strong>Parent-direct orchestration</strong>
|
|
173
|
+
<p>Dùng <span class="inline-code">/piagent-orchestration</span> để xem trần một helper, lý do dispatch/skip, projected token saving, review lenses, Field Guide và writer policy. Helper mặc định off. Opt-in vẫn chỉ cho một helper read-only context fresh khi chứng minh hai lane độc lập và tiết kiệm từ 30%; worker và retry bị tắt.</p>
|
|
174
174
|
</div>
|
|
175
175
|
|
|
176
176
|
<div class="command-grid">
|
|
@@ -29,7 +29,7 @@
|
|
|
29
29
|
},
|
|
30
30
|
"disjointness": {
|
|
31
31
|
"taxonomyDigest": "d42458d04861a60488a5f98532936db71d43dcccc0298d75b09c2255f2922cb1",
|
|
32
|
-
"publicExposureDigest": "
|
|
32
|
+
"publicExposureDigest": "80bed5d7ffd39dd9ebbc3207d5e6f09e9e56b9b254fd55814e6225677cb727f5",
|
|
33
33
|
"reportDigest": "5555555555555555555555555555555555555555555555555555555555555555",
|
|
34
34
|
"familyDisjoint": true,
|
|
35
35
|
"repositoryDisjoint": false,
|
|
@@ -29,7 +29,7 @@
|
|
|
29
29
|
},
|
|
30
30
|
"disjointness": {
|
|
31
31
|
"taxonomyDigest": "d42458d04861a60488a5f98532936db71d43dcccc0298d75b09c2255f2922cb1",
|
|
32
|
-
"publicExposureDigest": "
|
|
32
|
+
"publicExposureDigest": "80bed5d7ffd39dd9ebbc3207d5e6f09e9e56b9b254fd55814e6225677cb727f5",
|
|
33
33
|
"reportDigest": "5555555555555555555555555555555555555555555555555555555555555555",
|
|
34
34
|
"familyDisjoint": true,
|
|
35
35
|
"repositoryDisjoint": true,
|
|
@@ -0,0 +1,94 @@
|
|
|
1
|
+
{
|
|
2
|
+
"schemaVersion": 1,
|
|
3
|
+
"policy": {
|
|
4
|
+
"schemaVersion": 1,
|
|
5
|
+
"id": "codex-relative-efficiency-v1",
|
|
6
|
+
"baselineSurface": "codex-cli",
|
|
7
|
+
"candidateSurface": "piagent",
|
|
8
|
+
"estimand": "fixed-workload-family-geometric-mean",
|
|
9
|
+
"outcomeConditioning": "none",
|
|
10
|
+
"attemptPolicy": "accepted-plus-exact-provider-started-failed-attempts",
|
|
11
|
+
"maximumTotalTokenTrafficRatio": 0.7,
|
|
12
|
+
"maximumTotalTokenTrafficRatioUpper95": 0.7,
|
|
13
|
+
"maximumApiEquivalentCostRatio": 0.7,
|
|
14
|
+
"maximumApiEquivalentCostRatioUpper95": 0.7,
|
|
15
|
+
"maximumSubagentSessionsPerAttempt": 1,
|
|
16
|
+
"maximumSubagentTrafficShare": 0.05
|
|
17
|
+
},
|
|
18
|
+
"required": true,
|
|
19
|
+
"claim": "Piagent uses at least 30% less total provider traffic and API-equivalent cost than Codex CLI on the same work.",
|
|
20
|
+
"billedOrSubscriptionCost": false,
|
|
21
|
+
"costBoundary": "API-equivalent text-token pricing; not OAuth subscription debiting",
|
|
22
|
+
"checks": { "same-workload": true, "quality-noninferior": true },
|
|
23
|
+
"failures": [],
|
|
24
|
+
"passed": true,
|
|
25
|
+
"families": [
|
|
26
|
+
{
|
|
27
|
+
"scenarioId": "backend-change",
|
|
28
|
+
"expectedAttempts": 3,
|
|
29
|
+
"pairedAttempts": 3,
|
|
30
|
+
"exact": true,
|
|
31
|
+
"issues": [],
|
|
32
|
+
"baselineTokenTraffic": 300000,
|
|
33
|
+
"candidateTokenTraffic": 180000,
|
|
34
|
+
"totalTokenTrafficRatio": 0.6,
|
|
35
|
+
"baselineApiEquivalentCostUsd": 0.3,
|
|
36
|
+
"candidateApiEquivalentCostUsd": 0.18,
|
|
37
|
+
"apiEquivalentCostRatio": 0.6,
|
|
38
|
+
"candidateSubagentSessions": 0,
|
|
39
|
+
"candidateSubagentTraffic": 0,
|
|
40
|
+
"maximumSubagentSessionsPerAttempt": 0,
|
|
41
|
+
"candidateAttempts": 3,
|
|
42
|
+
"subagentEvidenceComplete": true
|
|
43
|
+
},
|
|
44
|
+
{
|
|
45
|
+
"scenarioId": "frontend-change",
|
|
46
|
+
"expectedAttempts": 3,
|
|
47
|
+
"pairedAttempts": 3,
|
|
48
|
+
"exact": true,
|
|
49
|
+
"issues": [],
|
|
50
|
+
"baselineTokenTraffic": 300000,
|
|
51
|
+
"candidateTokenTraffic": 180000,
|
|
52
|
+
"totalTokenTrafficRatio": 0.6,
|
|
53
|
+
"baselineApiEquivalentCostUsd": 0.3,
|
|
54
|
+
"candidateApiEquivalentCostUsd": 0.18,
|
|
55
|
+
"apiEquivalentCostRatio": 0.6,
|
|
56
|
+
"candidateSubagentSessions": 0,
|
|
57
|
+
"candidateSubagentTraffic": 0,
|
|
58
|
+
"maximumSubagentSessionsPerAttempt": 0,
|
|
59
|
+
"candidateAttempts": 3,
|
|
60
|
+
"subagentEvidenceComplete": true
|
|
61
|
+
}
|
|
62
|
+
],
|
|
63
|
+
"totalTokenTraffic": {
|
|
64
|
+
"complete": true,
|
|
65
|
+
"expectedScenarioFamilies": 2,
|
|
66
|
+
"completeScenarioFamilies": 2,
|
|
67
|
+
"ratio": 0.6,
|
|
68
|
+
"ratioConfidence95": { "lower": 0.6, "upper": 0.6, "sampleUnit": "scenario-family", "scenarioCount": 2 },
|
|
69
|
+
"ratioConfidence95Raw": { "lower": 0.6, "upper": 0.6, "sampleUnit": "scenario-family", "scenarioCount": 2 },
|
|
70
|
+
"aggregateBaselineTokens": 600000,
|
|
71
|
+
"aggregateCandidateTokens": 360000,
|
|
72
|
+
"aggregateRatio": 0.6,
|
|
73
|
+
"deltaPercent": -40
|
|
74
|
+
},
|
|
75
|
+
"apiEquivalentCost": {
|
|
76
|
+
"complete": true,
|
|
77
|
+
"expectedScenarioFamilies": 2,
|
|
78
|
+
"completeScenarioFamilies": 2,
|
|
79
|
+
"ratio": 0.6,
|
|
80
|
+
"ratioConfidence95": { "lower": 0.6, "upper": 0.6, "sampleUnit": "scenario-family", "scenarioCount": 2 },
|
|
81
|
+
"ratioConfidence95Raw": { "lower": 0.6, "upper": 0.6, "sampleUnit": "scenario-family", "scenarioCount": 2 },
|
|
82
|
+
"aggregateBaselineUsd": 0.6,
|
|
83
|
+
"aggregateCandidateUsd": 0.36,
|
|
84
|
+
"aggregateRatio": 0.6,
|
|
85
|
+
"deltaPercent": -40
|
|
86
|
+
},
|
|
87
|
+
"subagents": {
|
|
88
|
+
"candidateAttempts": 6,
|
|
89
|
+
"sessions": 0,
|
|
90
|
+
"tokenTraffic": 0,
|
|
91
|
+
"trafficShare": 0,
|
|
92
|
+
"maximumObservedSessionsPerAttempt": 0
|
|
93
|
+
}
|
|
94
|
+
}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"schemaVersion":
|
|
1
|
+
{"schemaVersion":2,"policyVersion":"role-policy-v1","role":"reviewer","objectiveHash":"aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa","objectiveText":"Review.","taskId":"task-1","taskRunId":"task-1-run-1","sessionHash":"bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb","authority":"read-only","readScope":["**"],"writeScope":["src/**"],"allowedTools":["read","edit"],"model":null,"contextBudget":6000,"ceilings":{"timeSeconds":600,"calls":40,"retries":0},"outputSchema":"review-v1","stoppingRule":"Stop.","approvalRestrictions":[],"parentAuthorityDigest":"cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc","singleWriterOwnership":null,"deduplicationKey":"dddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd","contextTransfer":{"mode":"isolated-minimal","inheritParentHistory":false,"maxSeedTokens":2048,"estimatedSeedTokens":101}}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"schemaVersion":
|
|
1
|
+
{"schemaVersion":2,"policyVersion":"role-policy-v1","role":"scout","objectiveHash":"aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa","objectiveText":"Map the bounded source entry points.","taskId":"task-1","taskRunId":"task-1-run-1","sessionHash":"bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb","authority":"read-only","readScope":["src/**"],"writeScope":[],"allowedTools":["read","grep","find","ls"],"model":null,"contextBudget":6000,"ceilings":{"timeSeconds":600,"calls":40,"retries":0},"outputSchema":"scout-result-v1","stoppingRule":"Stop when bounded evidence is sufficient.","approvalRestrictions":["no-external-write"],"parentAuthorityDigest":"cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc","singleWriterOwnership":null,"deduplicationKey":"dddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd","contextTransfer":{"mode":"isolated-minimal","inheritParentHistory":false,"maxSeedTokens":2048,"estimatedSeedTokens":125}}
|
|
@@ -16,6 +16,13 @@
|
|
|
16
16
|
"treeRevision": "subagent-tree.fixture",
|
|
17
17
|
"evidenceState": "complete",
|
|
18
18
|
"orchestration": { "mode": "bounded-subagents", "subagents": "used" },
|
|
19
|
+
"delegation": {
|
|
20
|
+
"action": "dispatch",
|
|
21
|
+
"reasonCodes": ["independent-lanes", "projected-net-saving"],
|
|
22
|
+
"projectedNetSavingsRatio": 0.35,
|
|
23
|
+
"minimumRequiredNetSavingsRatio": 0.3,
|
|
24
|
+
"budget": { "maxConcurrent": 1, "maxTotal": 1, "maxRetries": 0, "maxWriters": 0 }
|
|
25
|
+
},
|
|
19
26
|
"parent": { "nodeRef": "parent.fixture", "lifecycleState": "active", "budgetTerminal": false, "mergeOwner": "parent" },
|
|
20
27
|
"children": [
|
|
21
28
|
{
|