FEAT: Adding Temporary Instance Creation from the Registry

doc/code/framework.md

@@ -403,6 +403,8 @@ See [message normalizers](./targets/11_message_normalizer) for capability behavi
 
 - If you are creating a component with user input (e.g. via config, REST, or automatically) it should always use the registry
 - If you are storing an instance of a component, it should always use the registry
+- Construction does not require registration. The caller owns temporary instances
+  and their cleanup; the registry does not retain them.
 - The registry accepts only explicitly supported external inputs, permits opaque Python objects only for in-process callers, and leaves component validation to constructors.
 
 ## [Setup](./setup/0_setup)

doc/code/registry/0_registry.md

@@ -75,6 +75,62 @@ creation can build more than one component for the same name, but only one
 registration succeeds unless replacement is requested. The backend does not
 need a separate registration lock.
 
+### Construction without registration
+
+Use `create_instance(type_name, **params)` to build an object without adding it
+to `.instances`. The caller owns that object and its cleanup.
+`create_named_instance(...)` remains the build-and-register operation.
+
+The existing `POST /api/targets`, `POST /api/converters`, and `POST /api/scorers`
+endpoints accept `register: false`. Registration defaults to `true`, so existing
+clients keep their behavior. For example:
+
+```json
+{
+  "type": "CaesarConverter",
+  "register": false,
+  "params": { "caesar_offset": 3 }
+}
+```
+
+The response contains an `identifier` and, for targets, `capabilities`. It does
+not contain a registry name or a reusable object handle. The backend discards
+the built object after returning its descriptor.
+
+Targets and converters can also use a registered source that supports
+reconstruction:
+
+```json
+{
+  "type": "OpenAIChatTarget",
+  "register": false,
+  "source": {
+    "source_name": "objective",
+    "source_hash": "<source identifier hash>",
+    "params": { "temperature": 0.7 }
+  }
+}
+```
+
+Use `source.params` for overrides; do not also supply `params`. The backend
+keeps the source's other constructor inputs, including authentication, on the
+server. It does not change the source. Missing, changed, or ambiguous sources
+produce an error. If a source was renamed, a unique matching hash can resolve it.
+An optional `effective_hash` checks the reconstructed object's identity.
+Registered descriptors report `reconstructable` and, for targets,
+`supports_temperature_override`. Unsupported inputs can also report
+`reconstruction_error`.
+Constructor inputs that are not in the registry's build contract are not
+discarded silently: reconstruction is rejected. Such sources can still be used
+as registered objects.
+Reconstruction retains resolved defaults from the declared constructor chain,
+including explicitly forwarded parent parameters. Arguments generated inside a
+constructor are not treated as extra inputs from its caller.
+
+This REST flag applies to targets, converters, and scorers. `AttackRegistry`
+remains a class catalog, not a store of attack instances, and has no new REST
+endpoint.
+
 Constructor annotations define parameter metadata and coercion. Enum parameters
 accept member names or values. Types that inherit `StructuredParameterValue` declare their
 allowed variants through `get_registry_input_variants()`; the registry

doc/gui/0_gui.md

@@ -102,7 +102,7 @@ The Chat view is the primary workspace for running interactive attacks against c
 
 #### Sending Messages
 
-For a new chat, your default objective target is preselected if it is available. Click the target badge in the shared toolbar beside the label controls to open the target dropdown. If no target is selected, click **Select a target** in the same place. Your choice applies to this chat without changing the default. Saved chats keep their original target. An attack saved without a target uses this dropdown until its first send binds the selected target.
+For a new chat, your default objective target is preselected if it is available. Click the target badge in the chat ribbon below the common labels and above the objective to open the target dropdown. If no target is selected, click **Select a target** in the same place. Your choice applies to this chat without changing the default. Saved chats keep their original target. An attack saved without a target uses this dropdown until its first send binds the selected target.
 
 Clicking **Chat** while already in a new chat keeps its target and draft. Starting
 a new attack resets both. Default changes in another tab apply to the next new
@@ -112,6 +112,22 @@ Type a message and press Enter (or click Send) to send it to the chat target. Th
 
 When you open a saved chat, CoPyRIT automatically selects the target originally used, if its registered identity still matches. This also applies to direct links, reloads, and browser Back/Forward navigation. You can continue the same conversation without selecting the target again. Opening a saved chat does not change your defaults.
 
+#### Temperature
+
+Select a target, then set **Temperature** before the first send. Leave it empty
+to keep the source setting. The supported range is 0 to 2. This creates a private
+target configuration for the attack; it does not add or change a registered target.
+Temperature is read-only after the attack is bound. In the conversation editor,
+a temperature change requires **New attack**, not **Same attack**.
+Hover over or click the read-only field to see why it cannot be changed.
+
+Saved attacks retain the source name, source identity, temperature, and effective
+identity, but not credentials. After a restart, the backend tries to reconstruct
+the target from a matching registered source. If reconstruction fails, sending
+is blocked rather than using the source's default temperature. OpenAI-family
+targets with a temperature parameter support this control. Externally owned HTTP
+clients and temperature set through `extra_body_parameters` are not supported.
+
 #### Repeating a Message
 
 Use **n=1** beside Send to choose **1 to 10** repetitions. Enter and Send use the
@@ -161,6 +177,20 @@ messages endpoint and `send=false` context storage are unchanged.
 
 Open **Converters** and use the picker above the working input to add registered
 converters in the order you want them to run.
+For a stage that supports reconstruction, open **... > Settings** to change its
+constructor settings. Changes create a private converter for that stage. Other
+stages and the registered source keep their settings. **Reset to registered
+converter** removes the stage's overrides. **Use default / not set** clears a
+structured setting's override and restores the registered source's value.
+Changing settings invalidates that stage and its downstream results.
+
+Closing the converter pane discards temporary settings, but keeps content that
+you already applied and the identifiers of the converters that produced it.
+Temporary converters are built for each preview operation, not retained in a
+second registry. A runtime change clears temporary settings and preview results.
+Apply temporary converter results with **Add converted value** before repeating
+a message. Independent repeated conversion supports registered stages only.
+
 The top text box is an editable working copy: changing it does not change the original
 chat message. The top **Convert** button runs the active tab's configured pipeline
 and any configured inputs that do not have results yet. After every configured input
@@ -249,6 +279,15 @@ represents a manual conversion. `request_converter_configurations` controls
 conversion of pieces without a preconverted value; it does not describe which
 converters already ran.
 
+For temporary stages, preview requests provide `converter_specs`, aligned with
+`converter_ids`; use `null` for a registered stage. Preview responses include the
+actual temporary identifier and a signed `provenance` token. Pass these tokens
+as `applied_converter_provenance`, aligned with `applied_converter_ids`, when
+sending or saving the converted content. Tokens remain valid after the pane
+closes, without retaining converter objects. A restart invalidates tokens for
+unsaved content; convert it again. Stored messages retain their converter
+identifiers and do not depend on those tokens.
+
 #### Attachments
 
 Click the attachment button to add images, audio, video, or documents to your message. Supported types include `image/*`, `audio/*`, `video/*`, `.pdf`, `.doc`, `.docx`, and `.txt`. Attachments are displayed as chips below the input with type icons and file sizes.
@@ -371,7 +410,7 @@ Export stays available for read-only historical conversations, and is disabled w
 
 The labels bar above the page content is available across the GUI, including scanner setup, Home, Chat, and History. It shows the active labels for future attacks and scans, not the attribution of a historical run you are viewing. Click the labels icon to open **Default Labels** and add, edit, or remove custom labels. The required `operator` and `operation` controls remain in the bar, outside this popover, and cannot be removed. A signed-in operator is read-only.
 
-In Chat, the active target, Markdown toggle, export menu, conversations panel toggle, and **New Attack** button share the right side of this bar. They wrap below the labels on narrow screens.
+In Chat, a separate ribbon below the labels contains the target, temperature, and **Edit Conversation** controls on the left. The Markdown toggle, export menu, conversations panel toggle, and **New Attack** button are on the right.
 
 Clicking the `operation` label opens a picker listing the operations already recorded in memory, so you can choose one without typing it from memory. Typing a name that doesn't exist yet offers to create it. Very long lists show the first 200 and say how many are left, so type to narrow them. On narrow screens, use the labels icon to view or edit labels that do not fit inline.
 

frontend/src/App.labels.test.tsx

@@ -247,20 +247,34 @@ describe('Shared new run labels', () => {
     }))
   })
 
-  it('hosts Chat controls beside the labels and removes them when navigating away', async () => {
+  it('hosts Chat controls below the labels and above the objective, and removes them when navigating away', async () => {
     const user = userEvent.setup()
+    jest.mocked(targetsApi.listTargets).mockResolvedValue({
+      items: [{ ...TARGET, supports_temperature_override: true }],
+      pagination: { limit: 200, has_more: false },
+    })
     renderApp('/chat')
-    const toolbar = within(currentLabels()).getByRole('group', { name: 'Chat controls' })
-    const targetPicker = await within(toolbar).findByRole('combobox', { name: 'Chat target' })
+    const toolbar = screen.getByRole('group', { name: 'Chat controls' })
+    expect(currentLabels()).not.toContainElement(toolbar)
+    expect(currentLabels().compareDocumentPosition(toolbar) & Node.DOCUMENT_POSITION_FOLLOWING).toBeTruthy()
+    expect(toolbar.compareDocumentPosition(screen.getByTestId('objective-header')) & Node.DOCUMENT_POSITION_FOLLOWING)
+      .toBeTruthy()
+    const settings = within(toolbar).getByRole('group', { name: 'Conversation settings' })
+    const actions = within(toolbar).getByRole('group', { name: 'Conversation actions' })
+    const targetPicker = await within(settings).findByRole('combobox', { name: 'Chat target' })
     await waitFor(() => expect(targetPicker).toBeEnabled())
     await user.selectOptions(targetPicker, 'test_target')
     expect(targetPicker).toHaveValue('test_target')
+    const temperature = within(settings).getByRole('spinbutton', { name: 'Temperature' })
+    await user.type(temperature, '0.7')
+    expect(temperature).toHaveValue(0.7)
+    expect(within(settings).getByRole('button', { name: 'Edit Conversation' })).toBeEnabled()
     expect(within(screen.getByTestId('chat-area')).queryByRole('group', { name: 'Chat controls' }))
       .not.toBeInTheDocument()
-    expect(within(toolbar).getByRole('button', { name: 'Export conversation' })).toBeDisabled()
-    expect(within(toolbar).getByRole('button', { name: 'Toggle conversations panel' })).toBeDisabled()
-    expect(within(toolbar).getByRole('button', { name: 'New Attack' })).toBeDisabled()
-    const markdown = within(toolbar).getByRole('switch')
+    expect(within(actions).getByRole('button', { name: 'Export conversation' })).toBeDisabled()
+    expect(within(actions).getByRole('button', { name: 'Toggle conversations panel' })).toBeDisabled()
+    expect(within(actions).getByRole('button', { name: 'New Attack' })).toBeDisabled()
+    const markdown = within(actions).getByRole('switch')
     await user.click(markdown)
     expect(readUserPreferences('local').chatMarkdown).toBe(true)
 
@@ -271,7 +285,7 @@ describe('Shared new run labels', () => {
     expect(screen.queryByRole('group', { name: 'Chat controls' })).not.toBeInTheDocument()
     await user.click(screen.getByRole('button', { name: 'Chat', exact: true }))
     expect(screen.getAllByRole('group', { name: 'Chat controls' })).toHaveLength(1)
-    expect(within(currentLabels()).getByRole('switch')).toBeChecked()
+    expect(within(screen.getByRole('group', { name: 'Conversation actions' })).getByRole('switch')).toBeChecked()
   })
 
   it('preserves an edit made before the backend defaults arrive', async () => {
@@ -330,7 +344,7 @@ describe('Shared new run labels', () => {
     expect(attacksApi.createAttack).not.toHaveBeenCalled()
     expect(attacksApi.submitMessageSend).not.toHaveBeenCalled()
 
-    const toolbar = within(currentLabels()).getByRole('group', { name: 'Chat controls' })
+    const toolbar = screen.getByRole('group', { name: 'Chat controls' })
     expect(within(toolbar).getByLabelText('Active target: test_target')).toBeInTheDocument()
     await waitFor(() => expect(within(toolbar).getByRole('button', { name: 'Export conversation' })).toBeEnabled())
     await user.click(within(toolbar).getByRole('button', { name: 'Export conversation' }))

frontend/src/App.tsx

@@ -499,6 +499,7 @@ function AppContent({ operatorAlias }: { operatorAlias: string | null }) {
           endpoint: targetEndpoint(createdTarget),
           model_name: targetModelName(createdTarget),
           identifier_hash: targetIdentifierHash(createdTarget),
+          binding: createdTarget.binding,
         }
       : null
     skipNextLoadForAttackId.current = arId

frontend/src/components/Chat/ChatWindow.styles.ts

@@ -90,15 +90,34 @@ export const useChatWindowStyles = makeStyles({
     flexGrow: 1,
     flexWrap: 'wrap',
     gap: tokens.spacingHorizontalM,
+    minWidth: 0,
+    maxWidth: '100%',
+  },
+  conversationControls: {
+    display: 'flex',
+    alignItems: 'center',
+    flexWrap: 'wrap',
+    gap: tokens.spacingHorizontalM,
+    marginRight: 'auto',
+    minWidth: 0,
     maxWidth: '100%',
   },
   editActions: {
     display: 'flex',
     flexWrap: 'wrap',
     alignItems: 'center',
-    marginRight: 'auto',
     gap: tokens.spacingHorizontalXS,
   },
+  temperatureField: {
+    gridTemplateColumns: 'max-content 5rem',
+    columnGap: tokens.spacingHorizontalNone,
+    alignItems: 'center',
+    flexShrink: 0,
+  },
+  temperatureInput: {
+    width: '5rem',
+    minWidth: 0,
+  },
   sharedTarget: {
     maxWidth: '240px',
     [NARROW_VIEWPORT_QUERY]: {
@@ -123,6 +142,7 @@ export const useChatWindowStyles = makeStyles({
     flexWrap: 'wrap',
     justifyContent: 'flex-end',
     minWidth: 0,
+    marginLeft: 'auto',
   },
   ribbonAction: {
     ...mobileTouchTarget,