pi-agent-browser-native 0.1.6 → 0.2.0

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 CHANGED
@@ -1,5 +1,19 @@
1
1
  # Changelog
2
2
 
3
+ ## 0.2.0 - 2026-04-12
4
+
5
+ ### Changed
6
+ - `batch` now reuses the richer standalone renderers, so batched snapshots keep the compact main-content-first view and batched screenshots keep inline image attachments instead of degrading to raw JSON-ish text
7
+ - the tool schema now uses `sessionMode: "auto" | "fresh"` instead of the old implicit-session boolean so agents have a first-class way to request a fresh profiled/debug launch, and blocked startup-scoped reuse errors now include structured recovery hints
8
+ - plain-text inspection commands like `agent_browser --help` and `--version` are now always allowed, removing the old prompt-dependent inspection gate and making the inspection contract local and predictable
9
+ - navigation actions like `click`, `dblclick`, `back`, `forward`, and `reload` now include lightweight post-action title/url summaries when the wrapper can address the active session, reducing guess-and-check follow-up snapshots
10
+ - compact snapshot rendering is leaner by default: fewer additional sections, fewer refs, smaller role summaries, and the raw spill path now stays in `details.fullOutputPath` instead of dominating the visible snapshot body
11
+ - README and injected tool guidance now include a compact agent quick start with the core call shapes for `open` + `snapshot`, `click` + re-snapshot, `batch`, `eval --stdin`, and fresh profiled launches
12
+
13
+ ### Migration notes
14
+ - replace any use of `useActiveSession` with `sessionMode`
15
+ - use `sessionMode: "fresh"` when you need a new `--profile`, `--session-name`, or `--cdp` launch after the implicit session is already active
16
+
3
17
  ## 0.1.6 - 2026-04-12
4
18
 
5
19
  ### Changed
package/README.md CHANGED
@@ -87,6 +87,67 @@ This avoids duplicate `agent_browser` registrations if you also have the publish
87
87
 
88
88
  The native tool exposed to the agent is named `agent_browser`.
89
89
 
90
+ The primary session control parameter is `sessionMode`:
91
+
92
+ - `"auto"` (default) reuses the implicit `pi`-scoped session when possible
93
+ - `"fresh"` skips that implicit session so startup-scoped flags like `--profile`, `--session-name`, and `--cdp` can launch a fresh upstream session
94
+
95
+ ## Agent quick start
96
+
97
+ ### Mental model
98
+
99
+ - `args` — exact CLI args after `agent-browser`
100
+ - `stdin` — raw stdin only for `batch` and `eval --stdin`
101
+ - `sessionMode`
102
+ - `"auto"` — default, reuse the implicit `pi`-scoped session
103
+ - `"fresh"` — skip the implicit session for a new profile/debug launch
104
+
105
+ ### Common call shapes
106
+
107
+ Open a page, then take an interactive snapshot:
108
+
109
+ ```json
110
+ { "args": ["open", "https://example.com"] }
111
+ { "args": ["snapshot", "-i"] }
112
+ ```
113
+
114
+ Click a ref, then re-snapshot after navigation or a major DOM change:
115
+
116
+ ```json
117
+ { "args": ["click", "@e2"] }
118
+ { "args": ["snapshot", "-i"] }
119
+ ```
120
+
121
+ Run a multi-step browser flow in one tool call:
122
+
123
+ ```json
124
+ { "args": ["batch"], "stdin": "[[\"open\",\"https://example.com\"],[\"snapshot\",\"-i\"]]" }
125
+ ```
126
+
127
+ Evaluate page JavaScript via stdin:
128
+
129
+ ```json
130
+ { "args": ["eval", "--stdin"], "stdin": "document.title" }
131
+ ```
132
+
133
+ Start a fresh profiled launch after you already used the implicit session:
134
+
135
+ ```json
136
+ { "args": ["--profile", "Default", "open", "https://example.com/account"], "sessionMode": "fresh" }
137
+ ```
138
+
139
+ Name a new upstream session explicitly when you want to keep reusing it:
140
+
141
+ ```json
142
+ { "args": ["--session", "auth-flow", "open", "https://example.com"] }
143
+ ```
144
+
145
+ ### First useful prompt in a fresh `pi` session
146
+
147
+ ```text
148
+ Use the agent_browser tool to open https://react.dev and then take an interactive snapshot.
149
+ ```
150
+
90
151
  ## Local development
91
152
 
92
153
  Do not track or rely on a repo-local `.pi/extensions/agent-browser.ts` autoload shim for this package. When the package is also installed globally, that creates a duplicate `agent_browser` registration and blocks `pi` startup from this working directory.
@@ -116,14 +177,40 @@ Validated workflow examples:
116
177
  - run `batch` with JSON via `stdin`
117
178
  - run `eval --stdin`
118
179
  - take a screenshot with inline attachment support
119
- - inspect `agent_browser --help` and `--version`
180
+ - inspect `agent_browser --help` and `--version` via the tool's plain-text inspection fallback
181
+
182
+ Inspection commands like `agent_browser --help` and `--version` are always supported. They return plain text and are useful for debugging or capability checks, but they are not required for normal browsing workflows.
120
183
 
121
184
  Current cautions:
122
185
  - passing `--profile` is an explicit upstream choice; this extension does not add its own profile-cloning or isolation layer
123
- - startup-scoped flags like `--profile`, `--session-name`, and `--cdp` are for the first command that launches a session; if the implicit session is already active, the extension returns a validation error instead of silently letting upstream ignore those flags
186
+ - startup-scoped flags like `--profile`, `--session-name`, and `--cdp` are for the first command that launches a session; if the implicit session is already active, retry that call with `sessionMode: "fresh"` or provide an explicit `--session ...` for the new launch
124
187
  - implicit `piab-*` sessions are extension-managed convenience sessions; they are best-effort closed on `pi` shutdown, get an idle timeout to reduce stale background daemons, and clean up private temp spill artifacts on shutdown
125
188
  - explicit upstream sessions like `--session`, `--profile`, `--session-name`, and `--cdp` are treated as user-managed and are not auto-closed by the extension
126
189
 
190
+ ### Switching from public browsing to a fresh profile/debug launch
191
+
192
+ A common agent workflow is:
193
+
194
+ 1. browse a public page with the default implicit session
195
+ 2. then switch to a fresh authenticated/profile/debug launch
196
+
197
+ Use `sessionMode: "fresh"` for that transition instead of relying on the implicit session:
198
+
199
+ ```json
200
+ {
201
+ "args": ["--profile", "Default", "open", "https://example.com/account"],
202
+ "sessionMode": "fresh"
203
+ }
204
+ ```
205
+
206
+ If you want to name the new upstream session yourself, pass an explicit session instead:
207
+
208
+ ```json
209
+ {
210
+ "args": ["--session", "auth-flow", "--profile", "Default", "open", "https://example.com/account"]
211
+ }
212
+ ```
213
+
127
214
  ## Docs
128
215
 
129
216
  - [`docs/REQUIREMENTS.md`](docs/REQUIREMENTS.md) — product requirements and constraints
@@ -59,17 +59,19 @@ The published package should exclude agent-only and superseded repo materials su
59
59
 
60
60
  ### Default
61
61
 
62
- If the caller does not provide `--session`, the extension should use an implicit session name derived from the current `pi` session id.
62
+ If the caller does not provide `--session`, the extension should default to `sessionMode: "auto"` and use an implicit session name derived from the current `pi` session id.
63
63
 
64
64
  Why:
65
65
  - works out of the box
66
66
  - gives continuity across calls
67
67
  - avoids forcing the agent to invent session names for basic browsing
68
68
 
69
- ### Explicit upstream sessions
69
+ ### Explicit upstream sessions and fresh launches
70
70
 
71
71
  If the caller provides `--session`, `--profile`, `--cdp`, or similar upstream flags, the extension should respect them with minimal interference.
72
72
 
73
+ The tool should also expose a first-class `sessionMode: "fresh"` escape hatch so agents can intentionally skip the implicit session and launch a fresh upstream session without inventing a fixed explicit session name.
74
+
73
75
  ### Ownership
74
76
 
75
77
  V1 ownership rule:
@@ -92,7 +94,9 @@ The extension should surface that clearly and avoid hidden restart behavior in v
92
94
 
93
95
  That means explicit startup-scoping flags like `--profile`, `--session-name`, and `--cdp` should remain explicit upstream choices instead of being wrapped in extra hidden restart or cloning logic.
94
96
 
95
- If the implicit session is already active and one of those startup-scoped flags appears again, the extension should fail clearly instead of silently sending a command shape that upstream would ignore.
97
+ If the implicit session is already active and one of those startup-scoped flags appears again while `sessionMode` is still `"auto"`, the extension should fail clearly instead of silently sending a command shape that upstream would ignore.
98
+
99
+ That failure should include a structured recovery hint pointing to `sessionMode: "fresh"` as the first-line fix, while still allowing an explicit `--session` when the caller wants to name the new upstream session.
96
100
 
97
101
  ## Preferring the native tool
98
102
 
@@ -32,7 +32,7 @@ The tool also needs an operating playbook, not just a capability list. The model
32
32
  {
33
33
  "args": ["open", "https://example.com"],
34
34
  "stdin": "optional raw stdin content",
35
- "useActiveSession": true
35
+ "sessionMode": "auto"
36
36
  }
37
37
  ```
38
38
 
@@ -69,15 +69,20 @@ Examples:
69
69
  { "args": ["batch"], "stdin": "[[\"open\",\"https://example.com\"],[\"snapshot\",\"-i\"]]" }
70
70
  ```
71
71
 
72
- ### `useActiveSession`
72
+ ### `sessionMode`
73
73
 
74
- - type: `boolean`
74
+ - type: `"auto" | "fresh"`
75
75
  - optional
76
- - default: `true`
76
+ - default: `"auto"`
77
77
 
78
78
  Behavior:
79
79
  - if `args` already include `--session`, upstream session choice wins
80
- - otherwise the extension prepends its implicit active session when `useActiveSession` is `true`
80
+ - `"auto"` prepends the implicit active session when appropriate
81
+ - `"fresh"` skips the implicit session so startup-scoped flags like `--profile`, `--session-name`, or `--cdp` can launch a fresh upstream session
82
+
83
+ Recommended use:
84
+ - use `"auto"` for the common browse/snapshot/click flow inside one `pi` session
85
+ - use `"fresh"` when switching from an already-active implicit session to a new profile/debug/auth launch without inventing a fixed explicit session name
81
86
 
82
87
  ## Wrapper behavior
83
88
 
@@ -87,8 +92,8 @@ The extension should:
87
92
  - parse JSON output into tool details
88
93
  - handle observed JSON result shapes, including the array returned by `batch --json`
89
94
  - allow plain-text fallback for inspection commands like `--help` and `--version`
90
- - discourage exploratory inspection calls unless the user explicitly asks or debugging requires them
91
- - deflect normal-task `--help` inspection back into the standard browser workflow instead of letting the model relearn the tool from scratch each session
95
+ - support those inspection commands unconditionally so the tool contract stays local and predictable
96
+ - still describe normal browser workflows in guidance so models do not overuse inspection for routine tasks
92
97
  - surface stderr and non-zero exits clearly
93
98
  - attach images when the result points to a screenshot-like artifact
94
99
 
@@ -104,7 +109,8 @@ Primary content should be:
104
109
 
105
110
  Examples:
106
111
  - small `snapshot` results should include the actual snapshot text
107
- - oversized `snapshot` results should switch to a compact view that preserves the primary content, nearby sections, high-value refs, and a path to the spilled full raw snapshot
112
+ - oversized `snapshot` results should switch to a compact view that preserves the primary content, nearby sections, and a trimmed set of high-value refs, while exposing the full raw snapshot path via `details.fullOutputPath`
113
+ - successful navigation actions like `click`, `back`, `forward`, and `reload` should include a lightweight post-action title/url summary when the wrapper can address the active session
108
114
  - `tab list` should include a readable tab summary
109
115
  - `screenshot` should include the saved-path summary plus the inline image attachment when available
110
116
 
@@ -116,6 +122,7 @@ Recommended details:
116
122
  {
117
123
  "args": ["snapshot", "-i"],
118
124
  "effectiveArgs": ["--session", "pi-abc123", "--json", "snapshot", "-i"],
125
+ "sessionMode": "auto",
119
126
  "sessionName": "pi-abc123",
120
127
  "usedImplicitSession": true,
121
128
  "data": {
@@ -136,7 +143,8 @@ For oversized snapshots, details should switch to a compact metadata object and
136
143
 
137
144
  Worth doing in v1:
138
145
  - screenshots → inline image attachment
139
- - snapshots → origin + ref count + main-content-first compact preview, with full raw snapshot spill files when the inline result would otherwise be too large
146
+ - snapshots → origin + ref count + main-content-first compact preview, with the raw snapshot spill path kept in `details.fullOutputPath` when the inline result would otherwise be too large
147
+ - navigation actions like `click`, `back`, `forward`, and `reload` → lightweight post-action title/url summary when available
140
148
  - tab lists → compact summary/table
141
149
  - stream status → enabled/connected/port summary
142
150
 
@@ -158,7 +166,7 @@ If `agent-browser` is not on `PATH`, fail with a message that:
158
166
  - clean up private temp spill artifacts owned by the implicit session on shutdown
159
167
  - treat explicit upstream session choices like `--session`, `--profile`, `--session-name`, and `--cdp` as user-managed
160
168
  - pass explicit `--profile` straight through to upstream `agent-browser`; no profile-cloning or isolation layer is added in v1
161
- - if startup-scoped flags like `--profile`, `--session-name`, or `--cdp` are supplied after the implicit session is already active, return a validation error instead of silently relying on upstream to ignore them
169
+ - if startup-scoped flags like `--profile`, `--session-name`, or `--cdp` are supplied after the implicit session is already active while `sessionMode` is `"auto"`, return a validation error with a structured recovery hint that recommends `sessionMode: "fresh"`
162
170
 
163
171
  ## Non-goals
164
172
 
@@ -27,21 +27,29 @@ import {
27
27
  } from "./lib/runtime.js";
28
28
  import { cleanupSecureTempArtifacts } from "./lib/temp.js";
29
29
 
30
+ const DEFAULT_SESSION_MODE = "auto" as const;
31
+
30
32
  const AGENT_BROWSER_PARAMS = Type.Object({
31
33
  args: Type.Array(Type.String({ description: "Exact agent-browser CLI arguments, excluding the binary name." }), {
32
34
  description: "Exact agent-browser CLI arguments, excluding the binary name and any shell operators.",
33
35
  minItems: 1,
34
36
  }),
35
37
  stdin: Type.Optional(Type.String({ description: "Optional raw stdin content for commands like eval --stdin or batch." })),
36
- useActiveSession: Type.Optional(
37
- Type.Boolean({
38
- description: "When true and no explicit --session is present, inject the implicit session for this pi session.",
39
- default: true,
38
+ sessionMode: Type.Optional(
39
+ Type.Union([Type.Literal("auto"), Type.Literal("fresh")], {
40
+ description:
41
+ "Session handling mode. `auto` reuses the implicit pi-scoped session when possible. `fresh` skips the implicit session so startup-scoped flags like --profile, --session-name, or --cdp can launch a fresh upstream session.",
42
+ default: DEFAULT_SESSION_MODE,
40
43
  }),
41
44
  ),
42
45
  });
43
46
  const PROJECT_RULE_PROMPT =
44
47
  "Project rule: when browser automation is needed, prefer the native `agent_browser` tool. Do not run direct `agent-browser` bash commands unless the user explicitly asks for a bash-oriented workflow or browser-integration debugging.";
48
+ const QUICK_START_GUIDELINES = [
49
+ "Quick start mental model: args are the exact agent-browser CLI args after the binary; stdin is only for batch and eval --stdin; sessionMode=fresh starts a fresh upstream launch when you need new --profile, --session-name, or --cdp state.",
50
+ "Common first calls: { args: [\"open\", \"https://example.com\"] } then { args: [\"snapshot\", \"-i\"] }; after navigation, use { args: [\"click\", \"@e2\"] } then { args: [\"snapshot\", \"-i\"] }.",
51
+ "Common advanced calls: { args: [\"batch\"], stdin: \"[[\\\"open\\\",\\\"https://example.com\\\"],[\\\"snapshot\\\",\\\"-i\\\"]]\" }, { args: [\"eval\", \"--stdin\"], stdin: \"document.title\" }, and { args: [\"--profile\", \"Default\", \"open\", \"https://example.com/account\"], sessionMode: \"fresh\" }.",
52
+ ] as const;
45
53
  const BRAVE_SEARCH_PROMPT_GUIDELINE =
46
54
  "When a non-empty BRAVE_API_KEY is available in the current environment, prefer the Brave Search API via bash/curl to discover specific destination URLs, then open the chosen URL with agent_browser instead of browsing a search engine results page just to find the target.";
47
55
  const SHARED_BROWSER_PLAYBOOK_GUIDELINES = [
@@ -49,6 +57,7 @@ const SHARED_BROWSER_PLAYBOOK_GUIDELINES = [
49
57
  "For authenticated or user-specific content like feeds, inboxes, dashboards, and accounts, prefer --profile Default on the first browser call and let the implicit session carry continuity. Use --auto-connect only if profile-based reuse is unavailable or the task is specifically about attaching to a running debug-enabled browser.",
50
58
  "Do not invent fixed explicit session names for routine tasks. Use the implicit session unless you truly need multiple isolated browser sessions in the same conversation.",
51
59
  "When using --profile, --session-name, or --cdp, put them on the first command for that session. If you intentionally use an explicit --session, keep using that same explicit session for follow-ups.",
60
+ "If you already used the implicit session and now need startup-scoped flags like --profile, --session-name, or --cdp, retry with sessionMode set to fresh or pass an explicit --session for the new launch.",
52
61
  "If a session lands on the wrong page or tab, an interaction changes origin unexpectedly, or an open call returns blocked, blank, or otherwise unexpected results, use tab list / tab <n> / snapshot -i to recover state before retrying different URLs or fallback strategies. Only use wait with an explicit argument like milliseconds, --load, --url, --fn, or --text.",
53
62
  "For feed, timeline, or inbox reading tasks, focus on the main timeline/list region and read the first item there rather than unrelated composer or sidebar content.",
54
63
  "For read-only browsing tasks, prefer extracting the answer from the current snapshot, structured ref labels, or eval --stdin on the current page before navigating away. Only click into media viewers, detail routes, or new pages when the current view does not contain the needed information.",
@@ -62,7 +71,8 @@ const TOOL_PROMPT_GUIDELINES_SUFFIX = [
62
71
  "Do not fall back to osascript, AppleScript, or generic browser-driving bash commands when this tool can do the job.",
63
72
  "Pass exact agent-browser CLI arguments in args, excluding the binary name.",
64
73
  "Use stdin for commands like eval --stdin and batch instead of shell heredocs.",
65
- "Let the implicit session handle the common path unless you explicitly need upstream flags like --session, --profile, or --cdp.",
74
+ "Let the implicit session handle the common path unless you explicitly need a fresh launch for upstream flags like --profile, --session-name, or --cdp.",
75
+ "Use sessionMode=fresh when switching from an existing implicit session to a new profile/debug launch without inventing a fixed explicit session name.",
66
76
  ] as const;
67
77
 
68
78
  function buildMissingBinaryMessage(): string {
@@ -92,15 +102,81 @@ function isPlainTextInspectionArgs(args: string[]): boolean {
92
102
  return args.includes("--help") || args.includes("-h") || args.includes("--version") || args.includes("-V");
93
103
  }
94
104
 
95
- function buildInspectionDeflectionMessage(): string {
96
- return [
97
- "Do not inspect agent_browser help for a normal browser task.",
98
- "Use the workflow directly:",
99
- "1. open the target URL",
100
- "2. snapshot -i",
101
- "3. interact using refs and re-snapshot after navigation or major DOM changes",
102
- "For authenticated or user-specific content like feeds, inboxes, dashboards, or accounts, start with an authenticated strategy such as --profile Default on the first browser call and let the implicit session carry continuity. Use --auto-connect only if profile-based reuse is unavailable.",
103
- ].join("\n");
105
+ const NAVIGATION_SUMMARY_COMMANDS = new Set(["back", "click", "dblclick", "forward", "reload"]);
106
+
107
+ interface NavigationSummary {
108
+ title?: string;
109
+ url?: string;
110
+ }
111
+
112
+ function isRecord(value: unknown): value is Record<string, unknown> {
113
+ return typeof value === "object" && value !== null;
114
+ }
115
+
116
+ function shouldCaptureNavigationSummary(command: string | undefined, data: unknown): boolean {
117
+ return (
118
+ command !== undefined &&
119
+ NAVIGATION_SUMMARY_COMMANDS.has(command) &&
120
+ (!isRecord(data) || (typeof data.title !== "string" && typeof data.url !== "string"))
121
+ );
122
+ }
123
+
124
+ function extractStringResultField(data: unknown, fieldName: "title" | "url"): string | undefined {
125
+ if (typeof data === "string") {
126
+ const text = data.trim();
127
+ return text.length > 0 ? text : undefined;
128
+ }
129
+ if (!isRecord(data) || typeof data[fieldName] !== "string") {
130
+ return undefined;
131
+ }
132
+ const text = data[fieldName].trim();
133
+ return text.length > 0 ? text : undefined;
134
+ }
135
+
136
+ async function collectNavigationSummary(options: {
137
+ cwd: string;
138
+ sessionName?: string;
139
+ signal?: AbortSignal;
140
+ }): Promise<NavigationSummary | undefined> {
141
+ const { cwd, sessionName, signal } = options;
142
+ if (!sessionName) return undefined;
143
+
144
+ const readField = async (fieldName: "title" | "url"): Promise<string | undefined> => {
145
+ const processResult = await runAgentBrowserProcess({
146
+ args: ["--json", "--session", sessionName, "get", fieldName],
147
+ cwd,
148
+ signal,
149
+ });
150
+ if (processResult.aborted || processResult.spawnError || processResult.exitCode !== 0) {
151
+ return undefined;
152
+ }
153
+ const parsed = await parseAgentBrowserEnvelope({
154
+ stdout: processResult.stdout,
155
+ stdoutPath: processResult.stdoutSpillPath,
156
+ });
157
+ try {
158
+ if (parsed.parseError || parsed.envelope?.success === false) {
159
+ return undefined;
160
+ }
161
+ return extractStringResultField(parsed.envelope?.data, fieldName);
162
+ } finally {
163
+ if (processResult.stdoutSpillPath) {
164
+ await rm(processResult.stdoutSpillPath, { force: true }).catch(() => undefined);
165
+ }
166
+ }
167
+ };
168
+
169
+ const title = await readField("title");
170
+ const url = await readField("url");
171
+ if (!title && !url) return undefined;
172
+ return { title, url };
173
+ }
174
+
175
+ function mergeNavigationSummaryIntoData(data: unknown, navigationSummary: NavigationSummary): unknown {
176
+ if (isRecord(data)) {
177
+ return { ...data, navigationSummary };
178
+ }
179
+ return { navigationSummary, result: data };
104
180
  }
105
181
 
106
182
  function buildSharedBrowserPlaybookGuidelines(hasBraveApiKey: boolean): string[] {
@@ -115,6 +191,9 @@ function buildBrowserSystemPromptAppendix(hasBraveApiKey: boolean): string {
115
191
  return [
116
192
  PROJECT_RULE_PROMPT,
117
193
  "",
194
+ "Quick start:",
195
+ ...QUICK_START_GUIDELINES.map((guideline) => `- ${guideline}`),
196
+ "",
118
197
  "Browser operating playbook:",
119
198
  ...buildSharedBrowserPlaybookGuidelines(hasBraveApiKey).map((guideline) => `- ${guideline}`),
120
199
  ].join("\n");
@@ -123,6 +202,7 @@ function buildBrowserSystemPromptAppendix(hasBraveApiKey: boolean): string {
123
202
  function buildToolPromptGuidelines(hasBraveApiKey: boolean): string[] {
124
203
  return [
125
204
  ...TOOL_PROMPT_GUIDELINES_PREFIX,
205
+ ...QUICK_START_GUIDELINES,
126
206
  ...buildSharedBrowserPlaybookGuidelines(hasBraveApiKey),
127
207
  ...TOOL_PROMPT_GUIDELINES_SUFFIX,
128
208
  ];
@@ -194,16 +274,6 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
194
274
  promptGuidelines: toolPromptGuidelines,
195
275
  parameters: AGENT_BROWSER_PARAMS,
196
276
  async execute(_toolCallId, params, signal, onUpdate, ctx) {
197
- const promptPolicy = buildPromptPolicy(getLatestUserPrompt(ctx.sessionManager.getBranch()));
198
- if (!promptPolicy.allowAgentBrowserInspection && isPlainTextInspectionArgs(params.args)) {
199
- const errorText = buildInspectionDeflectionMessage();
200
- return {
201
- content: [{ type: "text", text: errorText }],
202
- details: { args: params.args, inspectionBlocked: true },
203
- isError: true,
204
- };
205
- }
206
-
207
277
  const validationError = validateToolArgs(params.args);
208
278
  if (validationError) {
209
279
  return {
@@ -213,10 +283,11 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
213
283
  };
214
284
  }
215
285
 
286
+ const sessionMode = params.sessionMode ?? DEFAULT_SESSION_MODE;
216
287
  const executionPlan = buildExecutionPlan(params.args, {
217
288
  implicitSessionActive,
218
289
  implicitSessionName,
219
- useActiveSession: params.useActiveSession ?? true,
290
+ sessionMode,
220
291
  });
221
292
 
222
293
  if (executionPlan.validationError) {
@@ -224,6 +295,8 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
224
295
  content: [{ type: "text", text: executionPlan.validationError }],
225
296
  details: {
226
297
  args: params.args,
298
+ sessionMode,
299
+ sessionRecoveryHint: executionPlan.recoveryHint,
227
300
  startupScopedFlags: executionPlan.startupScopedFlags,
228
301
  validationError: executionPlan.validationError,
229
302
  },
@@ -235,6 +308,7 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
235
308
  content: [{ type: "text", text: `Running agent-browser ${buildInvocationPreview(executionPlan.effectiveArgs)}` }],
236
309
  details: {
237
310
  effectiveArgs: executionPlan.effectiveArgs,
311
+ sessionMode,
238
312
  sessionName: executionPlan.sessionName,
239
313
  usedImplicitSession: executionPlan.usedImplicitSession,
240
314
  },
@@ -257,6 +331,7 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
257
331
  details: {
258
332
  args: params.args,
259
333
  effectiveArgs: executionPlan.effectiveArgs,
334
+ sessionMode,
260
335
  spawnError: processResult.spawnError.message,
261
336
  },
262
337
  isError: true,
@@ -268,12 +343,28 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
268
343
  stdout: processResult.stdout,
269
344
  stdoutPath: processResult.stdoutSpillPath,
270
345
  });
346
+ let presentationEnvelope = parsed.envelope;
271
347
  const processSucceeded = !processResult.aborted && !processResult.spawnError && processResult.exitCode === 0;
272
348
  const plainTextInspection = isPlainTextInspectionArgs(params.args) && processSucceeded && parsed.parseError !== undefined;
273
349
  const envelopeSuccess = plainTextInspection ? true : parsed.envelope?.success !== false;
274
350
  const parseSucceeded = plainTextInspection || parsed.parseError === undefined;
275
351
  const succeeded = processSucceeded && parseSucceeded && envelopeSuccess;
276
352
 
353
+ let navigationSummary: NavigationSummary | undefined;
354
+ if (succeeded && shouldCaptureNavigationSummary(executionPlan.commandInfo.command, parsed.envelope?.data)) {
355
+ navigationSummary = await collectNavigationSummary({
356
+ cwd: ctx.cwd,
357
+ sessionName: executionPlan.sessionName,
358
+ signal,
359
+ });
360
+ if (navigationSummary && presentationEnvelope) {
361
+ presentationEnvelope = {
362
+ ...presentationEnvelope,
363
+ data: mergeNavigationSummaryIntoData(presentationEnvelope.data, navigationSummary),
364
+ };
365
+ }
366
+ }
367
+
277
368
  implicitSessionActive = resolveImplicitSessionActiveState({
278
369
  command: executionPlan.commandInfo.command,
279
370
  priorActive: implicitSessionActive,
@@ -300,7 +391,7 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
300
391
  : await buildToolPresentation({
301
392
  commandInfo: executionPlan.commandInfo,
302
393
  cwd: ctx.cwd,
303
- envelope: parsed.envelope,
394
+ envelope: presentationEnvelope,
304
395
  errorText,
305
396
  });
306
397
 
@@ -308,16 +399,22 @@ export default function agentBrowserExtension(pi: ExtensionAPI) {
308
399
  content: presentation.content,
309
400
  details: {
310
401
  args: params.args,
402
+ batchSteps: presentation.batchSteps,
311
403
  command: executionPlan.commandInfo.command,
312
404
  subcommand: executionPlan.commandInfo.subcommand,
313
405
  data: presentation.data,
314
406
  error: parsed.envelope?.error,
407
+ navigationSummary,
315
408
  effectiveArgs: executionPlan.effectiveArgs,
316
409
  exitCode: processResult.exitCode,
317
410
  fullOutputPath: presentation.fullOutputPath,
411
+ fullOutputPaths: presentation.fullOutputPaths,
318
412
  imagePath: presentation.imagePath,
413
+ imagePaths: presentation.imagePaths,
319
414
  parseError: parsed.parseError,
415
+ sessionMode,
320
416
  sessionName: executionPlan.sessionName,
417
+ sessionRecoveryHint: executionPlan.recoveryHint,
321
418
  startupScopedFlags: executionPlan.startupScopedFlags,
322
419
  stderr: processResult.stderr || undefined,
323
420
  stdout: parseSucceeded ? undefined : processResult.stdout,
@@ -9,9 +9,17 @@
9
9
  import { readFile, stat } from "node:fs/promises";
10
10
  import { resolve } from "node:path";
11
11
 
12
- import type { CommandInfo } from "../runtime.js";
12
+ import { parseCommandInfo, type CommandInfo } from "../runtime.js";
13
13
  import { buildSnapshotPresentation, formatRawSnapshotText, formatSnapshotSummary } from "./snapshot.js";
14
- import { type AgentBrowserBatchResult, type AgentBrowserEnvelope, type ToolPresentation, isRecord, parsePositiveInteger, stringifyUnknown } from "./shared.js";
14
+ import {
15
+ type AgentBrowserBatchResult,
16
+ type AgentBrowserEnvelope,
17
+ type BatchStepPresentationDetails,
18
+ type ToolPresentation,
19
+ isRecord,
20
+ parsePositiveInteger,
21
+ stringifyUnknown,
22
+ } from "./shared.js";
15
23
 
16
24
  const IMAGE_EXTENSION_TO_MIME_TYPE: Record<string, string> = {
17
25
  ".gif": "image/gif",
@@ -23,6 +31,13 @@ const IMAGE_EXTENSION_TO_MIME_TYPE: Record<string, string> = {
23
31
 
24
32
  const INLINE_IMAGE_MAX_BYTES_ENV = "PI_AGENT_BROWSER_INLINE_IMAGE_MAX_BYTES";
25
33
  const DEFAULT_INLINE_IMAGE_MAX_BYTES = 5 * 1_024 * 1_024;
34
+ const NAVIGATION_SUMMARY_COMMANDS = new Set(["back", "click", "dblclick", "forward", "reload"]);
35
+ const NAVIGATION_SUMMARY_FIELD = "navigationSummary";
36
+
37
+ interface NavigationSummary {
38
+ title?: string;
39
+ url?: string;
40
+ }
26
41
 
27
42
  function getImageMimeType(filePath: string): string | undefined {
28
43
  const extension = filePath.toLowerCase().slice(filePath.lastIndexOf("."));
@@ -90,24 +105,213 @@ function getScreenshotSummary(data: Record<string, unknown>): string | undefined
90
105
  return typeof data.path === "string" ? `Saved image: ${data.path}` : undefined;
91
106
  }
92
107
 
93
- function formatBatchContent(data: AgentBrowserBatchResult[]): string {
94
- return data
95
- .map((item, index) => {
96
- const command = Array.isArray(item.command) ? item.command.join(" ") : `step-${index + 1}`;
97
- if (item.success === false) {
98
- return `${command}\nError: ${stringifyUnknown(item.error)}`;
99
- }
100
- return `${command}\n${stringifyUnknown(item.result)}`;
101
- })
108
+ function isNavigationObservableCommand(command: string | undefined): boolean {
109
+ return command !== undefined && NAVIGATION_SUMMARY_COMMANDS.has(command);
110
+ }
111
+
112
+ function isNavigationSummary(value: unknown): value is NavigationSummary {
113
+ return isRecord(value) && (typeof value.title === "string" || typeof value.url === "string");
114
+ }
115
+
116
+ function getNavigationSummary(data: Record<string, unknown>): NavigationSummary | undefined {
117
+ const candidate = data[NAVIGATION_SUMMARY_FIELD];
118
+ return isNavigationSummary(candidate) ? candidate : undefined;
119
+ }
120
+
121
+ function formatNavigationSummary(summary: NavigationSummary): string | undefined {
122
+ const title = typeof summary.title === "string" && summary.title.trim().length > 0 ? summary.title.trim() : undefined;
123
+ const url = typeof summary.url === "string" && summary.url.trim().length > 0 ? summary.url.trim() : undefined;
124
+ if (!title && !url) return undefined;
125
+ if (title && url) return `${title}\n${url}`;
126
+ return title ?? url;
127
+ }
128
+
129
+ function stripNavigationSummary(data: Record<string, unknown>): Record<string, unknown> {
130
+ const { [NAVIGATION_SUMMARY_FIELD]: _navigationSummary, ...rest } = data;
131
+ return rest;
132
+ }
133
+
134
+ function formatNavigationActionResult(data: Record<string, unknown>): string | undefined {
135
+ const actionData = stripNavigationSummary(data);
136
+ const lines: string[] = [];
137
+ if (typeof actionData.clicked === "string" || typeof actionData.clicked === "boolean") {
138
+ lines.push(`Clicked: ${String(actionData.clicked)}`);
139
+ }
140
+ if (typeof actionData.href === "string") {
141
+ lines.push(`Href: ${actionData.href}`);
142
+ }
143
+ if (typeof actionData.navigated === "boolean") {
144
+ lines.push(`Navigated: ${actionData.navigated}`);
145
+ }
146
+ if (lines.length > 0) {
147
+ return lines.join("\n");
148
+ }
149
+
150
+ const actionText = stringifyUnknown(actionData).trim();
151
+ if (actionText.length === 0 || actionText === "{}") {
152
+ return undefined;
153
+ }
154
+ return actionText;
155
+ }
156
+
157
+ function isStringArray(value: unknown): value is string[] {
158
+ return Array.isArray(value) && value.every((item) => typeof item === "string");
159
+ }
160
+
161
+ function getPresentationText(presentation: ToolPresentation): string {
162
+ return presentation.content
163
+ .filter((part): part is Extract<ToolPresentation["content"][number], { type: "text" }> => part.type === "text")
164
+ .map((part) => part.text.trim())
165
+ .filter((text) => text.length > 0)
102
166
  .join("\n\n");
103
167
  }
104
168
 
169
+ function getPresentationImages(presentation: ToolPresentation): Array<Extract<ToolPresentation["content"][number], { type: "image" }>> {
170
+ return presentation.content.filter(
171
+ (part): part is Extract<ToolPresentation["content"][number], { type: "image" }> => part.type === "image",
172
+ );
173
+ }
174
+
175
+ function getPresentationPaths(options: {
176
+ primaryPath?: string;
177
+ secondaryPaths?: string[];
178
+ }): string[] {
179
+ return options.secondaryPaths ?? (options.primaryPath ? [options.primaryPath] : []);
180
+ }
181
+
182
+ function formatBatchStepCommand(command: string[] | undefined, index: number): string {
183
+ return command && command.length > 0 ? command.join(" ") : `step-${index + 1}`;
184
+ }
185
+
186
+ function formatBatchStepError(error: unknown): string {
187
+ const errorText = stringifyUnknown(error).trim();
188
+ return errorText.length > 0 ? `Error: ${errorText}` : "Error: batch step failed.";
189
+ }
190
+
191
+ async function buildBatchStepPresentation(options: {
192
+ cwd: string;
193
+ index: number;
194
+ item: AgentBrowserBatchResult;
195
+ }): Promise<{ details: BatchStepPresentationDetails; presentation: ToolPresentation }> {
196
+ const { cwd, index, item } = options;
197
+ const command = isStringArray(item.command) ? item.command : undefined;
198
+ const commandText = formatBatchStepCommand(command, index);
199
+
200
+ if (item.success === false) {
201
+ const errorText = formatBatchStepError(item.error);
202
+ const presentation: ToolPresentation = {
203
+ content: [{ type: "text", text: errorText }],
204
+ summary: errorText,
205
+ };
206
+ return {
207
+ details: {
208
+ command,
209
+ commandText,
210
+ data: item.error,
211
+ index,
212
+ success: false,
213
+ summary: errorText,
214
+ text: errorText,
215
+ },
216
+ presentation,
217
+ };
218
+ }
219
+
220
+ const presentation = await buildToolPresentation({
221
+ commandInfo: parseCommandInfo(command ?? []),
222
+ cwd,
223
+ envelope: { data: item.result, success: true },
224
+ });
225
+ const fullOutputPaths = getPresentationPaths({
226
+ primaryPath: presentation.fullOutputPath,
227
+ secondaryPaths: presentation.fullOutputPaths,
228
+ });
229
+ const imagePaths = getPresentationPaths({
230
+ primaryPath: presentation.imagePath,
231
+ secondaryPaths: presentation.imagePaths,
232
+ });
233
+ const text = getPresentationText(presentation) || presentation.summary;
234
+
235
+ return {
236
+ details: {
237
+ command,
238
+ commandText,
239
+ data: presentation.data,
240
+ fullOutputPath: fullOutputPaths[0],
241
+ fullOutputPaths: fullOutputPaths.length > 0 ? fullOutputPaths : undefined,
242
+ imagePath: imagePaths[0],
243
+ imagePaths: imagePaths.length > 0 ? imagePaths : undefined,
244
+ index,
245
+ success: true,
246
+ summary: presentation.summary,
247
+ text,
248
+ },
249
+ presentation,
250
+ };
251
+ }
252
+
253
+ async function buildBatchPresentation(options: {
254
+ cwd: string;
255
+ data: AgentBrowserBatchResult[];
256
+ summary: string;
257
+ }): Promise<ToolPresentation> {
258
+ const { cwd, data, summary } = options;
259
+ const steps: Array<{ details: BatchStepPresentationDetails; presentation: ToolPresentation }> = [];
260
+ for (const [index, item] of data.entries()) {
261
+ steps.push(await buildBatchStepPresentation({ cwd, index, item }));
262
+ }
263
+
264
+ const images = steps.flatMap((step) => getPresentationImages(step.presentation));
265
+ const fullOutputPaths = steps.flatMap((step) => getPresentationPaths({
266
+ primaryPath: step.presentation.fullOutputPath,
267
+ secondaryPaths: step.presentation.fullOutputPaths,
268
+ }));
269
+ const imagePaths = steps.flatMap((step) => getPresentationPaths({
270
+ primaryPath: step.presentation.imagePath,
271
+ secondaryPaths: step.presentation.imagePaths,
272
+ }));
273
+ const text =
274
+ steps.length === 0
275
+ ? "(no batch steps)"
276
+ : steps
277
+ .map(({ details, presentation }) => {
278
+ const inlineImageCount = getPresentationImages(presentation).length;
279
+ const lines = [`Step ${details.index + 1} — ${details.commandText}`];
280
+ if (details.text.length > 0) {
281
+ lines.push(details.text);
282
+ }
283
+ if (inlineImageCount > 0) {
284
+ lines.push(`(${inlineImageCount} inline image attachment${inlineImageCount === 1 ? "" : "s"} below)`);
285
+ }
286
+ return lines.join("\n");
287
+ })
288
+ .join("\n\n");
289
+
290
+ return {
291
+ batchSteps: steps.map((step) => step.details),
292
+ content: [{ type: "text", text }, ...images],
293
+ data,
294
+ fullOutputPath: fullOutputPaths[0],
295
+ fullOutputPaths: fullOutputPaths.length > 0 ? fullOutputPaths : undefined,
296
+ imagePath: imagePaths[0],
297
+ imagePaths: imagePaths.length > 0 ? imagePaths : undefined,
298
+ summary,
299
+ };
300
+ }
301
+
105
302
  function formatSummary(commandInfo: CommandInfo, data: unknown): string {
106
303
  if (Array.isArray(data) && commandInfo.command === "batch") {
107
304
  const successCount = data.filter((item) => isRecord(item) && item.success !== false).length;
108
305
  return `Batch: ${successCount}/${data.length} succeeded`;
109
306
  }
110
307
  if (isRecord(data)) {
308
+ const navigationSummary = getNavigationSummary(data);
309
+ if (navigationSummary && isNavigationObservableCommand(commandInfo.command)) {
310
+ const navigationText = formatNavigationSummary(navigationSummary);
311
+ if (navigationText) {
312
+ return `${commandInfo.command ?? "navigation"} → ${navigationText.split("\n", 1)[0] ?? navigationText}`;
313
+ }
314
+ }
111
315
  if (commandInfo.command === "snapshot") {
112
316
  return formatSnapshotSummary(data);
113
317
  }
@@ -136,9 +340,6 @@ function formatSummary(commandInfo: CommandInfo, data: unknown): string {
136
340
  }
137
341
 
138
342
  function formatContentText(commandInfo: CommandInfo, data: unknown): string {
139
- if (Array.isArray(data) && commandInfo.command === "batch") {
140
- return formatBatchContent(data as AgentBrowserBatchResult[]);
141
- }
142
343
  if (typeof data === "string") {
143
344
  return data;
144
345
  }
@@ -149,6 +350,15 @@ function formatContentText(commandInfo: CommandInfo, data: unknown): string {
149
350
  return stringifyUnknown(data);
150
351
  }
151
352
 
353
+ const navigationSummary = getNavigationSummary(data);
354
+ if (navigationSummary && isNavigationObservableCommand(commandInfo.command)) {
355
+ const navigationText = formatNavigationSummary(navigationSummary);
356
+ if (navigationText) {
357
+ const actionText = formatNavigationActionResult(data);
358
+ return actionText ? `${actionText}\n\nCurrent page:\n${navigationText}` : `Current page:\n${navigationText}`;
359
+ }
360
+ }
361
+
152
362
  if (commandInfo.command === "snapshot") {
153
363
  return formatRawSnapshotText(data);
154
364
  }
@@ -232,13 +442,15 @@ export async function buildToolPresentation(options: {
232
442
  const data = envelope?.data;
233
443
  const summary = formatSummary(commandInfo, data);
234
444
  const presentation =
235
- commandInfo.command === "snapshot" && isRecord(data)
236
- ? await buildSnapshotPresentation(data)
237
- : {
238
- content: [{ type: "text" as const, text: formatContentText(commandInfo, data) }],
239
- data,
240
- summary,
241
- };
445
+ commandInfo.command === "batch" && Array.isArray(data)
446
+ ? await buildBatchPresentation({ cwd, data: data as AgentBrowserBatchResult[], summary })
447
+ : commandInfo.command === "snapshot" && isRecord(data)
448
+ ? await buildSnapshotPresentation(data)
449
+ : {
450
+ content: [{ type: "text" as const, text: formatContentText(commandInfo, data) }],
451
+ data,
452
+ summary,
453
+ };
242
454
 
243
455
  const imagePath = extractImagePath(cwd, data);
244
456
  if (!imagePath) {
@@ -19,11 +19,28 @@ export interface AgentBrowserBatchResult {
19
19
  success?: boolean;
20
20
  }
21
21
 
22
+ export interface BatchStepPresentationDetails {
23
+ command?: string[];
24
+ commandText: string;
25
+ data?: unknown;
26
+ fullOutputPath?: string;
27
+ fullOutputPaths?: string[];
28
+ imagePath?: string;
29
+ imagePaths?: string[];
30
+ index: number;
31
+ success: boolean;
32
+ summary: string;
33
+ text: string;
34
+ }
35
+
22
36
  export interface ToolPresentation {
37
+ batchSteps?: BatchStepPresentationDetails[];
23
38
  content: Array<{ text: string; type: "text" } | { data: string; mimeType: string; type: "image" }>;
24
39
  data?: unknown;
25
40
  fullOutputPath?: string;
41
+ fullOutputPaths?: string[];
26
42
  imagePath?: string;
43
+ imagePaths?: string[];
27
44
  summary: string;
28
45
  }
29
46
 
@@ -12,11 +12,13 @@ import { type ToolPresentation, compareRefIds, countLines, isRecord, normalizeWh
12
12
  const SNAPSHOT_INLINE_MAX_CHARS = 6_000;
13
13
  const SNAPSHOT_INLINE_MAX_LINES = 80;
14
14
  const SNAPSHOT_INLINE_MAX_REFS = 60;
15
- const SNAPSHOT_PRIMARY_PREVIEW_LINES = 10;
16
- const SNAPSHOT_SECTION_PREVIEW_LINES = 4;
17
- const SNAPSHOT_MAX_ADDITIONAL_SECTIONS = 5;
18
- const SNAPSHOT_KEY_REF_MAX_LINES = 20;
19
- const SNAPSHOT_OTHER_REF_MAX_LINES = 12;
15
+ const SNAPSHOT_PRIMARY_PREVIEW_LINES = 8;
16
+ const SNAPSHOT_SECTION_PREVIEW_LINES = 2;
17
+ const SNAPSHOT_MAX_ADDITIONAL_SECTIONS = 2;
18
+ const SNAPSHOT_KEY_REF_MAX_LINES = 8;
19
+ const SNAPSHOT_OTHER_REF_MAX_LINES = 4;
20
+ const SNAPSHOT_ROLE_COUNT_MAX_ENTRIES = 4;
21
+ const SNAPSHOT_FALLBACK_PREVIEW_MAX_LINES = 12;
20
22
  const SNAPSHOT_NAME_MAX_CHARS = 96;
21
23
  const SNAPSHOT_LINE_MAX_CHARS = 140;
22
24
  const SNAPSHOT_SPILL_FILE_PREFIX = "pi-agent-browser-snapshot";
@@ -165,7 +167,12 @@ function formatRoleCounts(roleCounts: Record<string, number>): string | undefine
165
167
  if (right[1] !== left[1]) return right[1] - left[1];
166
168
  return getRolePriority(left[0]) - getRolePriority(right[0]);
167
169
  });
168
- return ordered.map(([role, count]) => `${role} ${count}`).join(", ");
170
+ const visibleEntries = ordered.slice(0, SNAPSHOT_ROLE_COUNT_MAX_ENTRIES).map(([role, count]) => `${role} ${count}`);
171
+ const omittedEntries = Math.max(0, ordered.length - visibleEntries.length);
172
+ if (omittedEntries > 0) {
173
+ visibleEntries.push(`+${omittedEntries} more`);
174
+ }
175
+ return visibleEntries.join(", ");
169
176
  }
170
177
 
171
178
  function parseSnapshotLines(snapshot: string): SnapshotLine[] {
@@ -373,10 +380,10 @@ function buildSegmentPreview(segment: SnapshotSegment, maxLines: number): Snapsh
373
380
 
374
381
  function buildFallbackSnapshotOutline(snapshotLines: SnapshotLine[]): SnapshotPreview {
375
382
  const selected = new Set<number>();
376
- for (let index = 0; index < snapshotLines.length && selected.size < 6; index += 1) {
383
+ for (let index = 0; index < snapshotLines.length && selected.size < 4; index += 1) {
377
384
  if (!isNoiseSnapshotLine(snapshotLines[index])) selected.add(index);
378
385
  }
379
- for (let index = 0; index < snapshotLines.length && selected.size < 18; index += 1) {
386
+ for (let index = 0; index < snapshotLines.length && selected.size < SNAPSHOT_FALLBACK_PREVIEW_MAX_LINES; index += 1) {
380
387
  const line = snapshotLines[index];
381
388
  if (isNoiseSnapshotLine(line)) continue;
382
389
  if (SNAPSHOT_SIGNAL_ROLES.has(line.role) || line.ref || line.name.length > 0) {
@@ -385,7 +392,7 @@ function buildFallbackSnapshotOutline(snapshotLines: SnapshotLine[]): SnapshotPr
385
392
  }
386
393
  const chosenLines = [...selected]
387
394
  .sort((left, right) => left - right)
388
- .slice(0, 18)
395
+ .slice(0, SNAPSHOT_FALLBACK_PREVIEW_MAX_LINES)
389
396
  .map((index) => snapshotLines[index]);
390
397
  return {
391
398
  omittedCount: Math.max(0, snapshotLines.length - chosenLines.length),
@@ -508,6 +515,8 @@ export async function buildSnapshotPresentation(data: Record<string, unknown>):
508
515
  const snapshotSegments = useStructuredPreview ? buildSnapshotSegments(snapshotLines) : [];
509
516
  const primarySegment = useStructuredPreview ? choosePrimarySegment(snapshotSegments) : undefined;
510
517
  const additionalSegments = useStructuredPreview ? chooseAdditionalSegments(snapshotSegments, primarySegment) : [];
518
+ const additionalSegmentCount = useStructuredPreview && primarySegment ? Math.max(0, snapshotSegments.length - 1) : 0;
519
+ const omittedAdditionalSectionCount = Math.max(0, additionalSegmentCount - additionalSegments.length);
511
520
  const primaryPreview = primarySegment ? buildSegmentPreview(primarySegment, SNAPSHOT_PRIMARY_PREVIEW_LINES) : undefined;
512
521
  const additionalPreviews = additionalSegments
513
522
  .map((segment) => ({
@@ -548,11 +557,9 @@ export async function buildSnapshotPresentation(data: Record<string, unknown>):
548
557
  const lines: string[] = [
549
558
  `Origin: ${origin}`,
550
559
  `Refs: ${refEntries.length}`,
551
- ...(roleCountsText ? [`Roles: ${roleCountsText}`] : []),
560
+ ...(roleCountsText ? [`Top roles: ${roleCountsText}`] : []),
552
561
  "",
553
- fullOutputPath
554
- ? `Compact snapshot view. Full raw snapshot: ${fullOutputPath}`
555
- : `Compact snapshot view. Full raw snapshot unavailable: ${spillErrorText ?? "temp spill file could not be created."}`,
562
+ "Compact snapshot view.",
556
563
  ];
557
564
 
558
565
  if (fallbackPreview) {
@@ -581,6 +588,9 @@ export async function buildSnapshotPresentation(data: Record<string, unknown>):
581
588
  lines.push(`- ... (${preview.omittedCount} more lines in this section)`);
582
589
  }
583
590
  });
591
+ if (omittedAdditionalSectionCount > 0) {
592
+ lines.push(`- ... (${omittedAdditionalSectionCount} more sections omitted)`);
593
+ }
584
594
  }
585
595
  }
586
596
 
@@ -589,11 +599,16 @@ export async function buildSnapshotPresentation(data: Record<string, unknown>):
589
599
  lines.push("", "Other refs:", ...otherRefEntries.map(formatCompactRef));
590
600
  }
591
601
  if (omittedOtherRefs > 0) {
592
- lines.push(
593
- `- ... (${omittedOtherRefs} additional refs ${fullOutputPath ? "in the full snapshot file" : "were omitted with the full raw snapshot"})`,
594
- );
602
+ lines.push(`- ... (${omittedOtherRefs} additional refs omitted)`);
595
603
  }
596
604
 
605
+ lines.push(
606
+ "",
607
+ fullOutputPath
608
+ ? "Full raw snapshot path is available in details.fullOutputPath."
609
+ : `Full raw snapshot unavailable: ${spillErrorText ?? "temp spill file could not be created."}`,
610
+ );
611
+
597
612
  return {
598
613
  content: [{ type: "text", text: lines.join("\n") }],
599
614
  data: {
@@ -603,6 +618,7 @@ export async function buildSnapshotPresentation(data: Record<string, unknown>):
603
618
  previewMode: fallbackPreview ? "outline" : "structured",
604
619
  spillError: spillErrorText,
605
620
  previewRefIds: [...previewRefIds],
621
+ additionalSectionsOmitted: omittedAdditionalSectionCount,
606
622
  previewSections: [
607
623
  ...(primarySegment
608
624
  ? [
@@ -16,13 +16,6 @@ const IMPLICIT_SESSION_IDLE_TIMEOUT_ENV = "PI_AGENT_BROWSER_IMPLICIT_SESSION_IDL
16
16
  const IMPLICIT_SESSION_CLOSE_TIMEOUT_ENV = "PI_AGENT_BROWSER_IMPLICIT_SESSION_CLOSE_TIMEOUT_MS";
17
17
  const DEFAULT_IMPLICIT_SESSION_IDLE_TIMEOUT_MS = 15 * 60 * 1000;
18
18
  const DEFAULT_IMPLICIT_SESSION_CLOSE_TIMEOUT_MS = 5_000;
19
- const INSPECTION_ALLOW_PATTERNS = [
20
- /\bagent[_ -]?browser\s+--(?:help|version)\b/i,
21
- /\bagent[_ -]?browser\b.*\b(?:help|version|docs?|documentation|tool contract|tool guidance|tool description)\b/i,
22
- /\b(?:help|version|docs?|documentation|tool contract|tool guidance|tool description)\b.*\bagent[_ -]?browser\b/i,
23
- /\bdebug(?:ging)?\b.*\b(?:agent[_ -]?browser|agent_browser|browser integration)\b/i,
24
- /\bwhy\s+(?:isn't|is not|doesn't|does not)\b.*\b(?:agent[_ -]?browser|agent_browser)\b/i,
25
- ];
26
19
  const LEGACY_BASH_ALLOW_PATTERNS = [
27
20
  /\b(?:bash-oriented workflow|bash workflow)\b/i,
28
21
  /\b(?:use|via|through|with)\s+bash\b/i,
@@ -62,9 +55,19 @@ export interface CommandInfo {
62
55
  subcommand?: string;
63
56
  }
64
57
 
58
+ export type SessionMode = "auto" | "fresh";
59
+
60
+ export interface SessionRecoveryHint {
61
+ exampleArgs: string[];
62
+ exampleParams: { args: string[]; sessionMode: "fresh" };
63
+ reason: string;
64
+ recommendedSessionMode: "fresh";
65
+ }
66
+
65
67
  export interface ExecutionPlan {
66
68
  commandInfo: CommandInfo;
67
69
  effectiveArgs: string[];
70
+ recoveryHint?: SessionRecoveryHint;
68
71
  sessionName?: string;
69
72
  startupScopedFlags: string[];
70
73
  usedImplicitSession: boolean;
@@ -72,7 +75,6 @@ export interface ExecutionPlan {
72
75
  }
73
76
 
74
77
  export interface PromptPolicy {
75
- allowAgentBrowserInspection: boolean;
76
78
  allowLegacyAgentBrowserBash: boolean;
77
79
  }
78
80
 
@@ -177,7 +179,6 @@ export function getStartupScopedFlags(args: string[]): string[] {
177
179
 
178
180
  export function buildPromptPolicy(prompt: string): PromptPolicy {
179
181
  return {
180
- allowAgentBrowserInspection: INSPECTION_ALLOW_PATTERNS.some((pattern) => pattern.test(prompt)),
181
182
  allowLegacyAgentBrowserBash: LEGACY_BASH_ALLOW_PATTERNS.some((pattern) => pattern.test(prompt)),
182
183
  };
183
184
  }
@@ -212,21 +213,29 @@ export function getLatestUserPrompt(branch: unknown[]): string {
212
213
 
213
214
  export function buildExecutionPlan(
214
215
  args: string[],
215
- options: { implicitSessionActive: boolean; implicitSessionName: string; useActiveSession: boolean },
216
+ options: { implicitSessionActive: boolean; implicitSessionName: string; sessionMode: SessionMode },
216
217
  ): ExecutionPlan {
217
218
  const commandInfo = parseCommandInfo(args);
218
219
  const explicitSessionName = extractExplicitSessionName(args);
219
220
  const startupScopedFlags = getStartupScopedFlags(args);
220
221
  const effectiveArgs = args.includes("--json") ? [] : ["--json"];
222
+ let recoveryHint: SessionRecoveryHint | undefined;
221
223
  let sessionName = explicitSessionName;
222
224
  let usedImplicitSession = false;
223
225
  let validationError: string | undefined;
224
226
 
225
- if (!explicitSessionName && options.useActiveSession) {
227
+ if (!explicitSessionName && options.sessionMode === "auto") {
226
228
  if (options.implicitSessionActive && startupScopedFlags.length > 0) {
229
+ recoveryHint = {
230
+ exampleArgs: args,
231
+ exampleParams: { args, sessionMode: "fresh" },
232
+ reason:
233
+ "Startup-scoped flags like --profile, --session-name, and --cdp need a fresh upstream launch once the implicit session is already active.",
234
+ recommendedSessionMode: "fresh",
235
+ };
227
236
  validationError = [
228
237
  `The current implicit agent-browser session is already running, so startup-scoped flags ${startupScopedFlags.join(", ")} would be ignored by upstream agent-browser.`,
229
- "Reuse the existing implicit session without those flags, or start a fresh upstream session explicitly with `--session ...` (or `useActiveSession: false`) for a new launch.",
238
+ "Retry this call with `sessionMode: \"fresh\"` to force a fresh upstream launch, or pass an explicit `--session ...` if you want to name the new session yourself.",
230
239
  ].join(" ");
231
240
  } else {
232
241
  effectiveArgs.push("--session", options.implicitSessionName);
@@ -240,6 +249,7 @@ export function buildExecutionPlan(
240
249
  return {
241
250
  commandInfo,
242
251
  effectiveArgs,
252
+ recoveryHint,
243
253
  sessionName,
244
254
  startupScopedFlags,
245
255
  usedImplicitSession,
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "pi-agent-browser-native",
3
- "version": "0.1.6",
3
+ "version": "0.2.0",
4
4
  "description": "pi extension that exposes agent-browser as a native tool for browser automation",
5
5
  "type": "module",
6
6
  "author": "Mitch Fultz (https://github.com/fitchmultz)",