@raultov/browser-control-mcp 0.4.0-beta.7 → 0.6.0-beta.9

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
@@ -10,6 +10,136 @@ from a copy of its code: entries describe the differences from that release.
10
10
 
11
11
  ## [Unreleased]
12
12
 
13
+ ## [0.6.0-beta.9] - 2026-10-31
14
+
15
+ Public **BETA**. Spends fewer tokens and answers what the agent needs, after
16
+ a side-by-side study with Playwright MCP and chrome-devtools-mcp (now in the
17
+ README): `get_network_logs` and `get_console_logs` answer one line per entry
18
+ (details on demand), `snapshot` no longer repeats table text, input actions
19
+ wait for the page to settle, and `profile_page_performance` explains its
20
+ numbers with `insights`. Behaviour changes: the two log tools answer text
21
+ instead of JSON unless `include_details: true`, layout tables leave only
22
+ their content in `snapshot`, and input actions take ~50 ms more when the
23
+ page does not react.
24
+
25
+ ### Added
26
+
27
+ - `profile_page_performance` takes `insights: true`: what explains the
28
+ numbers, read in the page with the Web Performance APIs on both browsers:
29
+ the document's TTFB, server time and compression, the LCP element with
30
+ its breakdown (TTFB, load delay, load, render delay), layout-shift
31
+ culprits, render-blocking requests, the slowest requests and the scripts
32
+ of long animation frames (at most 5 each), and `findings` from the web.dev
33
+ thresholds (LCP over 2.5 s, CLS over 0.1, a slow server, an uncompressed
34
+ document, render-blocking requests). Firefox reports no layout shifts,
35
+ render-blocking status or long frames: those are `null`, with notes. The
36
+ reply grows from ~60 to ~600-700 tokens on Wikipedia. (#22)
37
+
38
+ ### Changed
39
+
40
+ - `click_element`, `hover`, `fill_input`, `fill_form` and `press_key` wait
41
+ for the page to settle before answering: until the DOM stays unchanged for
42
+ 50 ms (style changes aside) and the requests the action started finished,
43
+ at most 1 s, on both browsers. They answered in 1-2 ms, before the page
44
+ reacted, so a read right after a click that fetches data saw the old page.
45
+ An action nothing reacts to now takes ~50 ms and keeps its reply; one the
46
+ page reacts to adds how (`The page reacted (the DOM changed, 1 request(s))
47
+ and settled after 134 ms.`), or that it was still changing after 1 s. Not
48
+ done when the action left a dialog open or the page paused. (#21)
49
+ - `snapshot` and `find_in_snapshot` no longer print a table's text two or
50
+ three times. A table without headers (`th`, `thead`, `caption`,
51
+ `summary`) and without a role is layout: its `table`, `row` and `cell`
52
+ lines are left out and only their content stays, as in Chrome's
53
+ accessibility tree. A cell is named by its content only when that content
54
+ is plain text (folded into the name, up to 100 characters); a cell with
55
+ links or controls keeps them as children, unnamed. Hacker News went from
56
+ ~8 500 to ~4 500 tokens. (#20)
57
+ - `get_console_logs` answers one line per message by default: `[level]
58
+ text @ url:line` (lines from 1), repeats in a row folded (`[x3]`),
59
+ uncaught errors marked (`Uncaught ...`) with up to 5 stack lines
60
+ indented; `include_details: true` keeps the previous JSON. A page with
61
+ six messages went from ~550 to ~150 tokens. (#19)
62
+ - `get_network_logs` answers one line per request by default (`id METHOD
63
+ url => status [type]`, oldest first), then one per WebSocket frame with a
64
+ payload preview: ~100 tokens for a page of four requests instead of ~2 250.
65
+ The new `request_id` shows one request's headers and body;
66
+ `include_details: true` (now off by default) keeps the previous JSON with
67
+ every request's details. (#18)
68
+
69
+ ### Fixed
70
+
71
+ - `get_console_logs` previews objects and arrays as the console shows
72
+ them (`hello {n: 1, s: 'x'} [1, 2, 3]`): Chrome printed `Object`, Firefox
73
+ `Object(3) Array(3)`. On Chrome a `console.*` call keeps its location
74
+ (`url`, `line_number`), which was always empty, and `undefined`
75
+ arguments are printed. (#19)
76
+ - `get_network_logs` no longer lists the browser's own requests: a freshly
77
+ launched Chrome logged its new-tab page (`chrome://resources/...`, `data:`
78
+ images) before the first navigation. `chrome://`, `devtools://`, `data:`,
79
+ `about:` and extension URLs are not logged. (#18)
80
+
81
+ ## [0.5.0-beta.8] - 2026-10-31
82
+
83
+ Public **BETA**. Closes the rest of the Playwright MCP comparison: the new
84
+ `network` tool group mocks requests (`mock_route`, `list_routes`,
85
+ `remove_route`) and takes a tab offline (`set_offline`), `save_pdf` prints
86
+ a tab to PDF, and `--secrets` keeps known values out of what the agent
87
+ reads. Debugging a paused page works better: `evaluate_js` answers at once
88
+ on Chrome, a pause names its script, and the input tools fail at once
89
+ naming the pause instead of reporting a click or a text that has not
90
+ happened (a behaviour change: they used to "succeed").
91
+
92
+ ### Added
93
+
94
+ - `--secrets <file>` (dotenv): every tool result and error shows the
95
+ file's values as `<secret:NAME>` (as is or escaped in JSON), and
96
+ `fill_input` / `fill_form` type the value of a `<secret:NAME>` they are
97
+ given, so the agent can fill a password without reading it. A
98
+ convenience, not a security boundary: images are not masked and a page
99
+ can transform a value. Values never reach the logs. (#14)
100
+ - `save_pdf`, in the `files` tool group: prints the tab to PDF
101
+ (`landscape`, `print_background`, `scale`, `page_ranges`) and writes it to
102
+ `path`, an absolute `.pdf` under the upload roots, answering `{path,
103
+ bytes, pages}`; without `path` the PDF comes back as an embedded
104
+ `application/pdf` resource (at most 5 MB). Chrome `Page.printToPDF`,
105
+ Firefox `browsingContext.print`. (#13)
106
+ - `mock_route`, `list_routes`, `remove_route` and `set_offline`, in the new
107
+ `network` tool group (`--tool-groups network`). A route answers the tab's
108
+ requests whose URL matches a glob with a mocked status, body, content
109
+ type and headers, without reaching the network; the latest matching
110
+ route wins and `list_routes` counts its hits. Chrome uses the `Fetch`
111
+ domain (not while `enable_proxy_auth` is active), Firefox a
112
+ `beforeRequestSent` network intercept. `set_offline` takes the tab
113
+ offline or back online (Chrome still serves its HTTP cache). (#12)
114
+
115
+ ### Fixed
116
+
117
+ - The input tools (`click_element`, `hover`, `fill_input`, `fill_form`,
118
+ `press_key`) fail at once, naming the pause, when the page is already
119
+ paused in the debugger, and send nothing: the page handles input events
120
+ only once it resumes. Chrome reported a click or a key as done, and
121
+ both browsers blamed the field ("rejected the text") or, on Firefox, a
122
+ covering `html` element. (#17)
123
+ - On Chrome `evaluate_js` answers at once while the page is paused in the
124
+ debugger: it waited 10 s and failed with "no result after 10 s", even
125
+ for `1 + 1`, because Chrome cannot await an evaluation while no
126
+ microtask runs. It now evaluates without awaiting, reads a settled promise from
127
+ its state, and fails at once on a pending one, saying to `resume` or use
128
+ `evaluate_on_call_frame`. (#16)
129
+ - On Chrome a pause's location names its script: `resume`, `step_over`,
130
+ the `{paused, location}` of `click_element` and `press_key`, and the
131
+ "The page paused in the debugger at ..." message of an interrupted call
132
+ gave `url: null`, because Chrome leaves the deprecated
133
+ `callFrames[].url` empty. The URL now comes from the tab's parsed
134
+ scripts, and the location also carries `script_id` (what
135
+ `set_breakpoint` takes). (#10)
136
+
137
+ ### Changed
138
+
139
+ - The package description (crates.io, npm, Homebrew `desc`) is 79
140
+ characters instead of 213, within Homebrew's 80, so `brew style` no
141
+ longer flags the formula's `desc` line as too long.
142
+
13
143
  ## [0.4.0-beta.7] - 2026-10-31
14
144
 
15
145
  Public **BETA**. Closes the gaps a comparison with Playwright MCP showed:
package/README.md CHANGED
@@ -11,12 +11,62 @@ It is the successor of
11
11
  from a copy of its code) and complements Playwright MCP or
12
12
  chrome-devtools-mcp with runtime JavaScript debugging on both browsers.
13
13
 
14
- > **Status: BETA (0.4.0-beta.7).** This version contains known bugs and differences
14
+ > **Status: BETA (0.6.0-beta.9).** This version contains known bugs and differences
15
15
  > between Chrome and Firefox (see [Limitations and TODO](#limitations-and-todo))
16
16
  > and parts that have not been verified (see [Untested](#untested)). Tool
17
17
  > replies may change in incompatible ways before 1.0. Please report problems
18
18
  > in the [issues](https://github.com/raultov/browser-control-mcp/issues).
19
19
 
20
+ 💬 Questions, ideas or feedback: join us on [Discord](https://discord.gg/TunAHQZ8K).
21
+
22
+ ## Comparison with Playwright MCP and chrome-devtools-mcp
23
+
24
+ Measured on 2026-10-31 with browser-control-mcp built from the current
25
+ sources, `@playwright/mcp` 0.0.83 and `chrome-devtools-mcp` 1.12.0, driven by
26
+ the same stdio client through the same tasks on Chrome 156 (headless,
27
+ Linux). Every outcome was checked in the page, not taken from the reply.
28
+ Tokens are estimated as characters / 4 for every server.
29
+
30
+ **What each one can do**
31
+
32
+ | | browser-control-mcp | Playwright MCP | chrome-devtools-mcp |
33
+ | --- | --- | --- | --- |
34
+ | JavaScript debugger (breakpoints, call frames, stepping, script search) | **✓ Chrome and Firefox** (pause on load: Chrome) | ✗ | ✗ |
35
+ | Firefox | **✓ next to Chrome, same tools** | ✓ one browser per server | ✗ |
36
+ | Several browsers and instances at once | **✓** | ✗ | ✗ |
37
+ | Selectors | CSS, snapshot refs, `host >>> inner` into shadow roots and iframes | CSS, snapshot refs | snapshot refs only |
38
+ | Snapshot of a long page | **bounded at 50 000 characters**, `find_in_snapshot` searches the rest | unbounded | unbounded |
39
+ | Waits for the page to react after an action | ✓ ~50 ms when nothing reacts | ✓ ~500 ms | ✓ ~200 ms |
40
+ | Clicking a covered or hidden element | **fails at once, naming the overlay** | fails after a 5 s timeout | reports success (the overlay gets the click) |
41
+ | Mock requests, offline mode | ✓ | ✓ | offline only |
42
+ | PDF | ✓ file or inline | ✓ file | ✗ |
43
+ | Cookies and web storage | ✓ values hidden unless asked | ✓ values shown | ✗ |
44
+ | Secrets masked in output (`--secrets`) | ✓ | ✓ | ✗ |
45
+ | Performance | vitals + insights (LCP breakdown, layout shifts, render-blocking, long scripts) | trace file | trace + insights, Lighthouse |
46
+ | Heap snapshots, CSS styles, screencast | ✗ | video, tracing | ✓ |
47
+ | Generates Playwright test code | ✗ | ✓ | ✗ |
48
+ | Telemetry | **none** | none found | usage statistics on by default; traced URLs sent to CrUX |
49
+ | Single native binary (no Node.js) | **✓** | ✗ | ✗ |
50
+
51
+ **Measured**
52
+
53
+ | Task | browser-control-mcp | Playwright MCP | chrome-devtools-mcp |
54
+ | --- | --- | --- | --- |
55
+ | Tool list in context, default tools | 6 040 tokens | 5 330 tokens | 7 770 tokens |
56
+ | Tool list in context, every opt-in tool on | 9 500 tokens | 12 040 tokens | not measured (optional categories) |
57
+ | Snapshot of Hacker News | **4 588 tokens** | 12 356 tokens | 6 869 tokens |
58
+ | Snapshot of Wikipedia "Web browser" | **12 643 tokens** (cut, searchable) | 34 349 tokens | 27 656 tokens |
59
+ | Snapshot of MDN `<input>` | **12 625 tokens** (cut, searchable) | 158 077 tokens | 76 654 tokens |
60
+ | Log in step by step (8 calls) | **298 tokens, 1.8 s** | 521 tokens, 2.4 s | 364 tokens, 2.6 s |
61
+ | Log in with `fill_form` (5 calls) | **321 tokens, 1.6 s** | 428 tokens, 1.9 s | 352 tokens, 2.0 s |
62
+ | Click, median | 52 ms | 531 ms | 234 ms |
63
+ | Evaluate JavaScript, median | **0 ms** | 504 ms | 204 ms |
64
+ | Navigate (local page), median | **13 ms** | 31 ms | 119 ms |
65
+ | Full-page screenshot of a 31 700 px page | cut at 16 384 px, with a note | 3.3 MB image in the reply | saved to a file |
66
+ | Navigation to an unknown host | error | error | success with the error in the text |
67
+ | Navigate, then read the console (log, warning, errors) | 249 tokens | 191 tokens | 117 tokens |
68
+ | Navigate, then read the network log (3 requests) | 168 tokens | 48 tokens | 54 tokens |
69
+
20
70
  ## Supported browsers
21
71
 
22
72
  | Browser | Protocol | Validated with |
@@ -37,10 +87,11 @@ tools are off by default and listed only with `--tool-groups`:
37
87
  | Group | Tools | Why it is opt-in |
38
88
  | --- | --- | --- |
39
89
  | `storage` | `cookies`, `web_storage` | reads and changes session tokens |
40
- | `files` | `upload_file` | hands a page files of the machine |
90
+ | `files` | `upload_file`, `save_pdf` | reads and writes files of the machine |
41
91
  | `raw` | `send_cdp_command`, `send_bidi_command`, `send_rdp_request`, `get_custom_events` | anything the protocol allows |
42
92
  | `perf` | `get_performance_metrics`, `profile_page_performance` | niche |
43
93
  | `emulation` | `emulate` | niche |
94
+ | `network` | `mock_route`, `list_routes`, `remove_route`, `set_offline` | changes what the page receives |
44
95
 
45
96
  `--tool-groups all` lists every group. The page, input, observation and
46
97
  debugger tools are always listed (33 tools), plus `enable_proxy_auth` with
@@ -71,8 +122,9 @@ see the note.
71
122
  | `emulate` | ✓ | partial | `emulation` group; `width`/`height`, `device_scale_factor`, `mobile` (touch), `color_scheme`, `reduced_motion`, `forced_colors`, `media` (`print`), `reset`; answers what the page now sees; Firefox: viewport and device pixel ratio only (BiDi has no media or touch emulation yet; those options fail naming them) |
72
123
  | `wait_for` | ✓ | ✓ | waits (checking every 100 ms, at most 30 s) until a visible text appears or is gone, a selector is `visible` / `hidden` / `attached` / `detached`, or the URL contains a text; a dialog or debugger pause ends the wait at once |
73
124
  | `inspect_dom` | ✓ | ✓ | includes open shadow roots and same-origin iframes (cross-origin ones are noted); at most 100 000 characters without `query` (cut in the page) |
74
- | `snapshot` | ✓ | ✓ | outline of the page: role, name, states, value and a ref (`e12`) per element, through open shadow roots and same-origin iframes; at most 50 000 characters (`selector` and `depth` narrow it); the same JavaScript on both browsers |
125
+ | `snapshot` | ✓ | ✓ | outline of the page: role, name, states, value and a ref (`e12`) per element, through open shadow roots and same-origin iframes; tables without headers (`th`, `thead`, `caption`) are layout and leave only their content, and a cell is named by its text only when it holds nothing else; at most 50 000 characters (`selector` and `depth` narrow it); the same JavaScript on both browsers |
75
126
  | `find_in_snapshot` | ✓ | ✓ | `text` (case-insensitive) or `regex` over the whole outline; matching lines under their ancestors, at most 20 |
127
+ | `save_pdf` | ✓ | ✓ | `files` group; prints the tab as the browser's print does (`landscape`, `print_background`, `scale` 0.1–2, `page_ranges` like `1-3, 5`); with `path` (an absolute `.pdf` under the upload roots) writes the file and answers `{path, bytes, pages}`, else returns the PDF as an embedded `application/pdf` resource (at most 5 MB); Chrome `Page.printToPDF`, Firefox `browsingContext.print` |
76
128
  | `capture_screenshot` | ✓ | ✓ | `format` `png`, `jpeg` (or `jpg`) or `webp`; `full_page` at most 16 384 px per side (a text note says when cut) |
77
129
  | `click_element` | ✓ | ✓ | scrolls the element into view; `button` (left/right/middle), `click_count` (2 = double click) and `modifiers`; fails without clicking when it is hidden or covered; `host >>> inner` selectors reach open shadow roots and same-origin iframes (also in `fill_input`, `scroll`, `wait_for`); a dialog or debugger pause the click causes is noted in a successful reply (`{dialog}` or `{paused, location}`) |
78
130
  | `fill_input` | ✓ | ✓ | picks `<select>` options; fails without typing on elements that cannot take the text; types key by key on both (key handlers run; newlines and tabs inserted as text), about 1 ms per character; `change` on leaving the field, or at the end when the window has no focus |
@@ -84,11 +136,15 @@ see the note.
84
136
  | `evaluate_js` | ✓ | ✓ | `value` at most 100 000 characters (`truncated: true`); Chrome stops a script after 8 s, Firefox cannot (a looping tab is closed with `close_tab`) |
85
137
  | `cookies` | ✓ | ✓ | `storage` group; `get` / `set` / `delete`: the cookies sent to the page (or `url`) or of a `domain`; values hidden unless `show_values: true` (Chrome `Network.*Cookie*`, Firefox BiDi `storage.*` in the tab's partition) |
86
138
  | `web_storage` | ✓ | ✓ | `storage` group; `get` / `set` / `delete` / `clear` on the page's `localStorage` or `sessionStorage`; values hidden unless `show_values: true` |
87
- | `get_console_logs` | ✓ | ✓ | a message keeps at most 10 000 characters; Firefox: page errors, CSS warnings and platform messages need the debugger port |
88
- | `get_network_logs` | ✓ | ✓ | a failed request has `error`, a binary body `response_body_encoding: "base64"`; bodies at most 64 000 characters, none over 2 MB; a WebSocket payload keeps 64 000 (`payload_len` is the whole); Firefox: WebSocket frames need the debugger port |
139
+ | `get_console_logs` | ✓ | ✓ | one line per message, oldest first: `[level] text @ url:line` (lines from 1), objects and arrays previewed as the console shows them (`{n: 1}`, `[1, 2, 3]`), repeats in a row folded (`[x3]`), uncaught errors marked with up to 5 stack lines indented; `include_details: true` gives the JSON with every field; a message keeps at most 10 000 characters; Firefox: page errors, CSS warnings and platform messages need the debugger port |
140
+ | `get_network_logs` | ✓ | ✓ | one line per request, oldest first (`id METHOD url => status [type]`, or `failed: <reason>` / `pending`), then one per WebSocket frame (payload preview); `request_id` shows one request's headers and body, `include_details: true` every request's as JSON (a binary body has `response_body_encoding: "base64"`; bodies at most 64 000 characters, none over 2 MB; a WebSocket payload keeps 64 000, `payload_len` is the whole); the browser's own requests (`chrome://`, `devtools://`, `data:`, `about:`, extensions) are not logged; Firefox: WebSocket frames need the debugger port |
141
+ | `mock_route` | ✓ | ✓ | `network` group; requests whose URL matches a glob (`*`, `?`, `\` escapes) get a mocked `status`, `body` (text, at most 1 MB), `content_type` and `headers` without reaching the network; the latest matching route wins; at most 50 routes per tab, kept until `remove_route` or the tab closes (Chrome `Fetch`, not while `enable_proxy_auth` is active; Firefox a `network` intercept) |
142
+ | `list_routes` | ✓ | ✓ | `network` group; the tab's routes with the `hits` each answered |
143
+ | `remove_route` | ✓ | ✓ | `network` group; one route by id, or all |
144
+ | `set_offline` | ✓ | ✓ | `network` group; requests fail as without a network and `navigator.onLine` is false, until set back (Chrome `Network.emulateNetworkConditions`, Firefox `emulation.setNetworkConditions`); mocked routes still answer; Chrome still serves what its HTTP cache holds, as an offline browser does, Firefox fails every request |
89
145
  | `get_custom_events` | ✓ | partial | `raw` group; `{events, total, truncated}`, the newest `limit` events; `clear` removes the returned ones; params over 64 000 characters of JSON are replaced by `{truncated, characters}`; Firefox: events of the modules subscribed with `send_bidi_command` |
90
146
  | `get_performance_metrics` | ✓ | partial | `perf` group; Firefox: `Nodes`, `Frames`, `Resources`, `TransferSize` from the Web Performance APIs (no Performance domain), plus a note |
91
- | `profile_page_performance` | ✓ | partial | `perf` group; Firefox: FCP, LCP, DCL and Load from the Web Performance APIs; no trace, so `total_blocking_time_ms` and `top_long_tasks` are `null` |
147
+ | `profile_page_performance` | ✓ | partial | `perf` group; `insights: true` adds what explains the numbers, read with the Web Performance APIs (same JavaScript on both): the document's TTFB, server time and compression, the LCP element and its breakdown (TTFB, load delay, load, render delay), layout-shift culprits, render-blocking requests, the slowest requests and the scripts of long animation frames (at most 5 each), plus `findings` from the web.dev thresholds; Firefox: FCP, LCP, DCL and Load; no trace, so `total_blocking_time_ms` and `top_long_tasks` are `null`, and its insights have no layout shifts, render-blocking requests or long frames (`null`, with notes) |
92
148
  | `set_breakpoint` | ✓ | partial | Firefox: by URL only (no `script_id` / `script_hash`) |
93
149
  | `remove_breakpoint` | ✓ | ✓ | |
94
150
  | `resume` | ✓ | ✓ | |
@@ -105,6 +161,16 @@ see the note.
105
161
  | `webmcp_get_invocation` | ✓ | ✗ | needs `--enable-webmcp` |
106
162
  | `webmcp_list_invocations` | ✓ | ✗ | needs `--enable-webmcp` |
107
163
 
164
+ After `click_element`, `hover`, `fill_input`, `fill_form` and `press_key`
165
+ the server waits for the page to settle before answering: until its DOM
166
+ stays unchanged for 50 ms (style changes, often animations, aside) and the
167
+ requests the action started finished, at most 1 s, on both browsers. An
168
+ action nothing reacts to costs those 50 ms and keeps its reply; otherwise
169
+ the reply adds `The page reacted (the DOM changed, 1 request(s)) and settled
170
+ after 134 ms.`, or says the page was still changing after 1 s. A click that
171
+ navigates waits for the new page's requests. Use `wait_for` for anything
172
+ slower (a timer, a long request).
173
+
108
174
  The JavaScript debugger tools on Firefox need its debugger (RDP) port: on by
109
175
  default for a launched Firefox, `--firefox-rdp-port` for an attached one. The
110
176
  same port gives Firefox's WebSocket frames and the console messages BiDi does
@@ -379,8 +445,9 @@ claude mcp add browser-control-mcp -- npx -y @raultov/browser-control-mcp
379
445
  | `--enable-automation` | off | shows the automation infobar | ignored; `list_instances` shows a warning |
380
446
  | `--user-profile` | off | your default profile (no `--user-data-dir`) | only with `--firefox-port` (attach to your running Firefox) |
381
447
  | `--allow-cookie-import` | off | exposes `copy_cookies` | capability error |
382
- | `--tool-groups <GROUPS>` | none | lists the opt-in tool groups, comma separated: `storage`, `files`, `raw`, `perf`, `emulation` or `all` (see [Tool groups](#tool-groups)) | same |
383
- | `--upload-root <DIR>` | working + temp dir | directory `upload_file` may read files from (repeatable; replaces the defaults) | same |
448
+ | `--secrets <FILE>` | none | dotenv file (`NAME=value`, at least 4 characters per value) whose values every tool result and error shows as `<secret:NAME>`; `fill_input` and `fill_form` type the value of a `<secret:NAME>` (see [Security notes](#security-notes)) | same |
449
+ | `--tool-groups <GROUPS>` | none | lists the opt-in tool groups, comma separated: `storage`, `files`, `raw`, `perf`, `emulation`, `network` or `all` (see [Tool groups](#tool-groups)) | same |
450
+ | `--upload-root <DIR>` | working + temp dir | directory `upload_file` may read files from and `save_pdf` may write PDFs to (repeatable; replaces the defaults) | same |
384
451
  | `--enable-webmcp` | off | WebMCP switches and `webmcp_*` tools | capability error |
385
452
  | `--proxy-server <URL>` | none | proxy for every instance; exposes `enable_proxy_auth` | `http://` and `socks5://` proxies |
386
453
  | `--proxy-username <USER>` | none | answers proxy (407) challenges | same (untested) |
@@ -441,7 +508,13 @@ Chrome and Firefox take the same calls:
441
508
  unfinished (`evaluate_js`, `fill_input`, `navigate`...) fails at once
442
509
  with that location.
443
510
  5. `evaluate_on_call_frame` reads locals of the paused frame; `step_over`
444
- moves one statement; `resume` continues.
511
+ moves one statement; `resume` continues. While paused, the input tools
512
+ (`click_element`, `hover`, `fill_input`, `fill_form`, `press_key`) fail
513
+ at once naming the pause and do nothing: the page would handle their
514
+ events only after `resume`. `evaluate_js` still answers
515
+ while paused, in the page's global scope. A promise it returns cannot
516
+ always settle meanwhile: Chrome reads a settled one and fails at once on
517
+ a pending one; Firefox settles it when its microtasks can run.
445
518
  6. `remove_breakpoint` with the id `set_breakpoint` returned.
446
519
 
447
520
  Both browsers answer the same replies, with 0-based lines everywhere:
@@ -450,7 +523,7 @@ Both browsers answer the same replies, with 0-based lines everywhere:
450
523
  | --- | --- |
451
524
  | `set_breakpoint` | `{"breakpoint_id": "bp-N", "url", "line_number", "column_number", "resolved"}`: where the browser bound it; `resolved` is false until a script with that URL loads |
452
525
  | `remove_breakpoint` | `{"removed": "bp-N"}`; an unknown id fails listing the tab's ids |
453
- | `step_over`, `resume` | `{"paused", "location": {"url", "line_number", "column_number"} \| null}`: where the thread stopped next (`step_over` waits up to 2 s; `resume` reports a pause only if one follows within 250 ms) |
526
+ | `step_over`, `resume` | `{"paused", "location": {"url", "script_id"?, "line_number", "column_number"} \| null}`: where the thread stopped next (`script_id` on Chrome, for `set_breakpoint`) (`step_over` waits up to 2 s; `resume` reports a pause only if one follows within 250 ms) |
454
527
  | `evaluate_on_call_frame` | the `evaluate_js` shape; an exception fails with `Evaluation threw: <what was thrown>` |
455
528
 
456
529
  Breakpoint ids are per tab and MCP-level (`bp-1`, `bp-2`...), like tab ids.
@@ -534,7 +607,7 @@ incompatible:
534
607
  - Firefox debugger: tabs are mapped to the debugger through Firefox's Remote
535
608
  Agent (verified on Firefox 157 only); a Firefox without it answers the
536
609
  debugger tools with an error instead of guessing a tab.
537
- - `get_network_logs` with response bodies has no total size limit (each
610
+ - `get_network_logs` with `include_details: true` has no total size limit (each
538
611
  Firefox body is capped at 2 MB): use `url_filter` on busy pages.
539
612
  - The Docker image is amd64 only.
540
613
 
@@ -558,6 +631,19 @@ incompatible:
558
631
  the upload roots (the server's working directory and the system temp
559
632
  directory, or the `--upload-root` directories), resolved so `..` and
560
633
  symlinks cannot leave them. Paths are never logged.
634
+ - `--secrets <file>` keeps known values (test passwords, API keys) out of
635
+ what the agent reads: every tool result and error (text, embedded text,
636
+ structured content) shows them as `<secret:NAME>`, also when escaped
637
+ inside JSON, and `fill_input` / `fill_form` type the value of a
638
+ `<secret:NAME>` they are given, so the agent can log in without seeing
639
+ the password. It is a convenience, not a security boundary: the page can
640
+ transform a value (encode it, split it) past the masking, images
641
+ (screenshots, PDFs) are not masked, and the server stops at start-up if
642
+ the file has a malformed line, a value shorter than 4 characters or a
643
+ name defined twice (the error names the line, never a value). Values
644
+ never reach the logs.
645
+ - `save_pdf` writes only `.pdf` files, under the same upload roots and
646
+ resolved the same way; it overwrites an existing `.pdf` there.
561
647
  - The proxy password is never logged and is given only to proxy (407)
562
648
  challenges, never to a site's own authentication.
563
649
  - The Firefox debugger port needs `devtools.chrome.enabled`: any local
@@ -19,7 +19,7 @@
19
19
  "hasInstallScript": true,
20
20
  "license": "MIT",
21
21
  "name": "@raultov/browser-control-mcp",
22
- "version": "0.4.0-beta.7"
22
+ "version": "0.6.0-beta.9"
23
23
  },
24
24
  "node_modules/detect-libc": {
25
25
  "engines": {
@@ -48,5 +48,5 @@
48
48
  }
49
49
  },
50
50
  "requires": true,
51
- "version": "0.4.0-beta.7"
51
+ "version": "0.6.0-beta.9"
52
52
  }
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "artifactDownloadUrls": [
3
- "https://github.com/raultov/browser-control-mcp/releases/download/v0.4.0-beta.7"
3
+ "https://github.com/raultov/browser-control-mcp/releases/download/v0.6.0-beta.9"
4
4
  ],
5
5
  "author": "raultov <raultov@protonmail.com>",
6
6
  "bin": {
@@ -9,7 +9,7 @@
9
9
  "dependencies": {
10
10
  "detect-libc": "^2.1.2"
11
11
  },
12
- "description": "MCP server to control and debug Chrome and Firefox at runtime: breakpoints, call frames and stepping over CDP and WebDriver BiDi/RDP, plus navigation, DOM, input, screenshots, console and network capture.",
12
+ "description": "MCP server to control and debug Chrome and Firefox, with JavaScript breakpoints",
13
13
  "devDependencies": {
14
14
  "prettier": "^3.8.3"
15
15
  },
@@ -92,7 +92,7 @@
92
92
  "zipExt": ".tar.xz"
93
93
  }
94
94
  },
95
- "version": "0.4.0-beta.7",
95
+ "version": "0.6.0-beta.9",
96
96
  "volta": {
97
97
  "node": "18.14.1",
98
98
  "npm": "9.5.0"