CSP3 requires a complete base64-value, so truncated and illegal nonce
and hash tokens were being reported as real sources.
Co-authored-by: Cursor <cursoragent@cursor.com>
parseCaptureJson retained the strings jsonc-parser 3.3.1 builds fragment by
fragment around each escape, which V8 keeps as ropes several times larger
than their text. Escape-heavy HAR captures therefore exhausted the 192 MiB
decoder heap from about 15 MB, well inside the 32 MiB input budget.
Each string token is now decoded with JSON.parse from its accepted source
span, which yields the same value as one flat string. jsonc-parser reports
malformed strings through onError before the value callback, so JSON.parse
only sees tokens it already accepted. HAR, qlog and Playwright trace
decoding share this parser.
Fixes#10132
A one-shot CLI failure tells the caller to set the launch environment for the next REA process. update_analysis_settings stays on the continuing session. Reused JavaScript workers also keep the limit source that selected them.
artifact_limits and web-network budgets report update_mode parameter because update_analysis_settings cannot change them. The legacy-name scan no longer reads docs, and the 2026-11-09 Ghidra receipts are restored to the fields they recorded.
Callers read update_mode to see whether update_analysis_settings can change a limit, and pending_restart only when a configured value is not active in this process.
* fix(process): revalidate ownership before signaling
* fix(process): retain cleanup ownership until completion
* fix(process): keep omitted live descendants unresolved
A non-empty process table that omits a retained live PID cannot authorize a signal or certify cleanup. An empty table can still describe a token-verified owner.
* fix(process): read omitted Darwin run tokens by PID
A live identity that ps omitted was still validated through Darwin
environment(), which rejects any PID absent from a second process table.
Ask the native reader for that PID directly so the launch token can bind.
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
* fix(browser-scenario): page captured text on Unicode boundaries
* fix(browser-scenario): retain popup events when open metadata exceeds its budget
* fix(browser-scenario): redact saved secrets across wire encodings
* fix(browser-scenario)!: record parsed request actions in evidence
BREAKING CHANGE: New captures with request actions that save nothing include save: [] in the recorded parameters. Their scenario and Evidence digests can differ from older captures of the same scenario. Read retained captures with their original recorded digests and use the newly recorded digest for new captures.
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
* test: focus assertions on behavioral evidence
Prune redundant and wording-coupled checks while retaining diagnostic, parser, recovery, and cleanup coverage.
* refactor: remove test-only wrappers and duplicate checks
Exercise production operations through test fixtures, remove unused adapters, and keep managed-member uncertainty guidance under its existing owner.
* test: focus process validation on diagnostic facts
* refactor: remove test-only runtime helpers
Move semantic evidence fixture setup out of runtime and remove unused Windows settle recovery helpers while checking configured defaults independently.
* test: focus artifact diagnostics on format facts
* test: verify retained cleanup guidance delivery
* fix(native): retain value trace source evidence
* fix(native): build trace error evidence only on failure
* test(mcp): expect unowned value trace refusal
* test(mcp): avoid freezing trace error wording
* test(native): keep value-trace stop coverage with source evidence
Main's stop and continuation cases now meet the trace's evidence-owner
precondition: direct calls supply a source id, and a completed MCP trace
runs in a session that can retain it.
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
* fix(browser)!: keep the Chromium sandbox for REA-launched browsers
Playwright appends --no-sandbox unless a launch passes
chromiumSandbox: true. Neither the browser scenario launch nor the native
module resolver passed it, so every REA-launched Chromium ran without its
sandbox and REA_BROWSER_NO_SANDBOX had no effect.
Both launch sites now take the sandbox decision from ChromiumSandbox.ts:
off on Android or when REA_BROWSER_NO_SANDBOX=true, on otherwise. When
Chromium reports that it cannot start its sandbox, the launch failure's
remediation names the opt-out.
Fixes#8722
BREAKING CHANGE: hosts where Chromium cannot create its sandbox, such as
root users and Linux containers without user namespaces or a setuid
sandbox, now fail to launch until REA_BROWSER_NO_SANDBOX=true is set.
* fix(browser): name the sandbox opt-out on any sandboxed launch failure
Chromium reports an unusable sandbox through several unrelated startup
messages, including a misconfigured setuid helper and a refused namespace
move, and Playwright recognizes only some of them. Matching individual
messages missed real failures, so a sandboxed launch failure now always
names REA_BROWSER_NO_SANDBOX=true as a conditional remediation and the
retained startup diagnostics show whether the sandbox was the cause.
* fix(browser): keep default recovery for resolver failures after launch
The module resolver's outer catch also covers context, navigation and
evaluation failures after the browser launched. Attaching the sandbox
remediation there replaced their default recovery guidance. Attach the
shared launch remediation only when a sandboxed launch itself fails.
* docs(browser): note the Android sandbox exception for module traces
Kaoru0822