chore: sync actions from gh-aw@v0.83.5 - #201
Merged
Merged
Conversation
pelikhan
marked this pull request as ready for review
July 29, 2026 10:12
Copilot stopped reviewing on behalf of
pelikhan due to an error
July 29, 2026 10:32
Contributor
There was a problem hiding this comment.
Pull request overview
Note
Copilot could not run the full agentic suite for this review because it was automatically requested on a bot-authored pull request. Request a review from Copilot under Reviewers to retry with the full agentic suite. Improved support for bot-authored pull requests is coming soon.
This PR improves workflow robustness and diagnostics across agent runs by adding better failure categorization/remediation (notably missing model pricing), enhancing safe-outputs with optional structured data validation/persistence, and tightening infra handling (gateway/tool discovery, watchdog timeouts, firewall domain summaries, and PR review pinning).
Changes:
- Add a dedicated “missing model pricing” failure context with optional live pricing lookup + remediation template, and wire it into failure issue/comment templates.
- Introduce structured
datasupport for safe-outputs (schema normalization + validation + rendering in summaries/previews). - Improve operational resilience: pin PR reviews to specific commits, detect empty safeoutputs gateway tool registration, normalize MCP config regressions, centralize watchdog timeout logic, and filter internal sidecar domains from blocked-domain reporting.
Show a summary per file
| File | Description |
|---|---|
| setup/sh/convert_gateway_config_gemini.sh | Clarifies Gemini gateway host/domain behavior under network isolation. |
| setup/setup.sh | Adds new JS helper to safe-outputs allowlist in setup. |
| setup/md/missing_model_pricing.md | New remediation template for missing model pricing failures. |
| setup/md/agent_failure_issue.md | Injects missing-model-pricing context into failure issue body. |
| setup/md/agent_failure_comment.md | Injects missing-model-pricing context into failure comment body. |
| setup/js/submit_pr_review.cjs | Adds support to pin PR reviews to a specific commit SHA via config. |
| setup/js/start_mcp_gateway.cjs | Normalizes known sink-visibility double-encoding regression in MCP config. |
| setup/js/safe_outputs_tools.json | Updates tool schemas/descriptions; starts introducing structured-data documentation. |
| setup/js/safe_outputs_handlers.cjs | Validates/enforces structured data enablement + schema validation at handler layer. |
| setup/js/safe_output_type_validator.cjs | Validates structured data, normalizes it, and persists it via body appending. |
| setup/js/safe_output_summary.cjs | Renders structured data into generated step summaries. |
| setup/js/render_template.cjs | Preserves leading newline when stripping template blocks to avoid line-joining. |
| setup/js/push_to_pull_request_branch.cjs | Refactors git auth extraheader override into shared helper. |
| setup/js/process_runner.cjs | Centralizes post-result watchdog timeout constants and resolver. |
| setup/js/pr_review_buffer.cjs | Implements commit pinning for PR reviews (trigger-time SHA / configured SHA). |
| setup/js/parse_firewall_logs.cjs | Filters internal sidecar hostnames from blocked-domain reporting/summaries. |
| setup/js/package.json | Bumps Copilot SDK and updates minimatch version constraint. |
| setup/js/mount_mcp_as_cli.cjs | Adds “gateway empty tools” flag and fails fast when safeoutputs tools/list is empty. |
| setup/js/models.json | Adds pricing entry for gemini-3.6-flash. |
| setup/js/mcp_scripts_validation.cjs | Exposes validateValueAgainstSchema for non-argument schema validation. |
| setup/js/log_parser_shared.cjs | Adds structured data preview rendering in safe-outputs log preview. |
| setup/js/handle_agent_failure.cjs | Adds missing-model-pricing detection/context, fetches pricing from models.dev, and improves titles/categories. |
| setup/js/git_auth_helpers.cjs | Introduces shared withGitHubHostToken helper for temporary extraheader overrides. |
| setup/js/firewall_blocked_domains.cjs | Filters internal sidecar hostnames from blocked-domain warnings shown externally. |
| setup/js/detect_agent_errors.cjs | Detects missing-model-pricing via stdio + firewall audit logs, outputs model name. |
| setup/js/data_schema_normalizer.cjs | New normalizer for structured-data schemas (shorthand + keyword validation). |
| setup/js/create_pull_request.cjs | Uses shared git auth helper and scopes git ops via cwd for correctness. |
| setup/js/create_pr_review_comment.cjs | Adds support to pin PR review comments to a commit SHA via config. |
| setup/js/copilot_harness.cjs | Reuses centralized watchdog timeout logic and refines retry/success suppression rules. |
| setup/js/collect_ndjson_output.cjs | Fails hard on missing outputs when gateway-empty flag exists; passes structured-data config into validation. |
| setup/js/codex_harness.cjs | Adds post-result watchdog + terminal safe-output detection to treat exit-hangs as success. |
| setup/js/apply_samples.cjs | Improves PR head ref resolution using args/inputs/config target and adds helpers. |
| setup/js/ai_credits_context.cjs | Expands audit/event JSONL path probing and extracts model name for unknown-model AI credits events. |
Review details
Tip
Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
- Files reviewed: 33/33 changed files
- Comments generated: 7
- Review effort level: Low
Comment on lines
+378
to
+383
| if (entry.data !== undefined) { | ||
| const toolConfig = getSafeOutputsToolConfig(config, type); | ||
| const dataEnabled = toolConfig?.data_enabled === true || (toolConfig?.data_schema && typeof toolConfig.data_schema === "object"); | ||
| if (!dataEnabled) { | ||
| return buildIntentErrorResponse(`${type} data is not enabled (set safe-outputs.data in workflow frontmatter)`); | ||
| } |
Comment on lines
+97
to
+113
| const requiredSet = new Set(Object.keys(normalizedProperties)); | ||
| if (node.required !== undefined) { | ||
| if (!Array.isArray(node.required)) { | ||
| throw new Error(`${path}.required: must be an array of strings`); | ||
| } | ||
| for (let i = 0; i < node.required.length; i++) { | ||
| const requiredName = node.required[i]; | ||
| if (typeof requiredName !== "string" || requiredName.trim().length === 0) { | ||
| throw new Error(`${path}.required[${i}]: must be a non-empty string`); | ||
| } | ||
| if (!Object.prototype.hasOwnProperty.call(normalizedProperties, requiredName)) { | ||
| throw new Error(`${path}.required[${i}]: unknown property "${requiredName}"`); | ||
| } | ||
| requiredSet.add(requiredName); | ||
| } | ||
| } | ||
| result.required = [...requiredSet].sort(); |
Comment on lines
5
to
15
| "inputSchema": { | ||
| "type": "object", | ||
| "required": ["title", "body"], | ||
| "$defs": { | ||
| "structured_data": { | ||
| "type": "object", | ||
| "description": "Optional structured data to carry machine-readable context through sanitization-safe channels. When provided, this object is preserved and appended to the body as fenced JSON.", | ||
| "additionalProperties": true | ||
| } | ||
| }, | ||
| "properties": { |
Comment on lines
+778
to
+787
| // If this safe-output type supports a body field, append structured data | ||
| // as fenced JSON so it survives body sanitization. | ||
| if (Object.prototype.hasOwnProperty.call(typeConfig.fields, "body")) { | ||
| const dataBlock = `${STRUCTURED_DATA_LABEL}\n\`\`\`json\n${dataJSON}\n\`\`\``; | ||
| if (typeof normalizedItem.body === "string" && normalizedItem.body.length > 0) { | ||
| normalizedItem.body = `${normalizedItem.body}\n\n${dataBlock}`; | ||
| } else { | ||
| normalizedItem.body = dataBlock; | ||
| } | ||
| } |
Comment on lines
108
to
+118
| function recoverSafeOutputsToolsIfNeeded(tools, core) { | ||
| if (tools.length > 0) { | ||
| return tools; | ||
| } | ||
| const fallbackPath = process.env.GH_AW_SAFE_OUTPUTS_TOOLS_PATH || `${RUNNER_TEMP}/gh-aw/safeoutputs/tools.json`; | ||
| const recovered = loadToolsFromJSONFile(fallbackPath, core); | ||
| if (recovered.length > 0) { | ||
| core.warning(` safeoutputs tools/list returned empty; recovered ${recovered.length} tool(s) from ${fallbackPath}`); | ||
| return recovered; | ||
| } | ||
| throw new Error(`safeoutputs tool schema is empty (tools/list returned 0 and fallback ${fallbackPath} is empty/missing). ` + `Failing fast to avoid agent runs without discoverable safe-output tools.`); | ||
|
|
||
| // The live MCP gateway returned 0 tools for safeoutputs. Write a flag file so | ||
| // that collect_ndjson_output.cjs can surface this as a hard failure instead of | ||
| // silently concluding "graceful no-op" when outputs.jsonl is never written. | ||
| writeSafeOutputsGatewayEmptyFlag(core); | ||
|
|
||
| throw new Error(`safeoutputs tools/list returned 0 tools. ` + `Failing fast — the live MCP gateway has no tools registered. ` + `Check the MCP gateway startup logs for ECONNRESET errors or delayed backend registration.`); |
| "@vitest/coverage-v8": "^4.1.8", | ||
| "@vitest/ui": "^4.1.10", | ||
| "minimatch": ">=3.1.3", | ||
| "minimatch": ">=10.2.6", |
Comment on lines
+1832
to
+1840
| /** | ||
| * Build a frontmatter YAML pricing snippet for the missing model. | ||
| * Returns null when pricing data is unavailable. | ||
| * @param {string} modelName | ||
| * @param {string} engineId | ||
| * @param {{input: number, output: number, cacheRead?: number, cacheWrite?: number}|null} pricing Per-million-token values from models.dev | ||
| * @returns {string|null} | ||
| */ | ||
| function buildModelPricingFrontmatterSnippet(modelName, engineId, pricing, isPlaceholderPricing = false) { |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Automated sync of actions from gh-aw at
v0.83.5.