v0.97.0

What's Changed

MR: #438
Full Changelog: v0.96.10...v0.97.0

Addresses:

• #437
• #441

Includes:

• task() method exported from cecli.main to run system programmatically as a part of another program
• /terminal-setup updates to allow the user to confirm configuration edits
• Fix JSON parsing from raw responses so weaker models can perform tool calling outside of the OpenAI tool calling schema
• Automatic command backgrounding so agent cannot get stuck on miscalled, non-exiting command
• New key bindings for TUI (input_start, input_end, output_up, and output_down)

v0.96.10

What's Changed

MR: #435
Full Changelog: v0.96.9...v0.96.10

Addresses:

• #288
• #408
• #409
• #417
• #432

Includes:

• Simplify line based editing tools (InsertText and DeleteText) to work off of pure line numbers
• Make Grep tool capable of multi-searching
• Speed up git ignored file checking by a factor of ~100
• Speed up file mention checking due to the above
• Small tool error mitigations

v0.96.9

What's Changed

MR: #431
Full Changelog: v0.96.8...v0.96.9

Includes:

• Fix for error on switching to gemini models from non-gemini model
• Fix for error when model duplicates tool call responses

v0.96.8

What's Changed

MR: #427
Full Changelog: v0.96.7...v0.96.8

Includes:

• Format system messages and in context files before adding user message to conversation stream
• Use relative paths consistently in conversation messages so models know we are referring to the same files

v0.96.7

What's Changed

MR: #424
Full Changelog: v0.96.6...v0.96.7

Addresses:

• #421

Includes:

• Add environment variable for cecli config file path by @gopar in #422
• Feature/rebrand website (remove Aider references, ensure site is cohesive with project identity) by @synth-mania in #423
• Allow ShowNumberedContext and ReplaceText tools to be used for multiple parallel searches and edits to cut down the number of exploration and edit turns required for models powerful enough to take advantage of it
• Change tool executor methods so models are allowed to hallucinate parameters without necessarily disrupting operations as long as they specify a proper superset of the real parameters
• Allow the /add command to include files outside of the current repository

New Contributors

• @synth-mania made their first contribution in #423

v0.96.6

What's Changed

MR: #420
Full Changelog: v0.96.5...v0.96.6

Addresses:
#419

Includes:

• TUI output rendering speed up for large outputs by removing reliance on Rich's console.capture()
• Better parsing of delimited commands

v0.96.5

What's Changed

MR: #418
Full Changelog: v0.96.4...v0.96.5

Includes:

• Autocomplete that functions in classic mode across all slash commands
• Autocomplete on empty prefix for initial recommendations
• Remind editor model that it actually needs to apply the edits from the architect model
• More (and forced) autosave checkpoints

v0.96.4

What's Changed

MR: #416
Full Changelog: v0.96.3...v0.96.4

Addresses:
#413

Includes:

• Stricter message ordering logic to preserve context

v0.96.3

What's Changed

MR: #412
Full Changelog: v0.96.2...v0.96.3

Includes:

• /compact {message} to force compaction with hints to what the model should focus on
• Fix chunk ordering on creation of temporary coder class creation for one-off messages (used extensively across the code base)

v0.96.2

What's Changed

MR: #411
Full Changelog: v0.96.1...v0.96.2

Includes:

• Upgrade litellm minimum version to better enforce their data contract
• Fix /tokens to count tokens using the conversation system proeprly
• thought_signatures error and translating them to thought_signature for general Gemini model operation
• Restore original SSE transport implementation since being a subclass of the Http based transport doesn't work with the MCP SDK's signature expectations