Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
Claims 1, 8 & 15 are amended. Claims 1-20 are presented for examination.
Response to Arguments
Applicant’s arguments filed on 4/22/2026 have been reviewed. Following are the responses to the amendments.
Rejection under 35 U.S.C. 101
Applicant arguments have been considered and are persuasive, hence the rejection under 35 U.S.C. 101 is withdrawn.
Rejection under 35 U.S.C. 102
Applicant’s arguments, with respect to the rejection(s) of claim(s) 1, 5, 8, 12, 15 & 19 under 102(a)(2) have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new grounds of rejection has been given over Enlaz (WO 2025141192 / US Provisional: 63/616,230) and further in view of Keller (US 20250156535 A1) and Tyagi (US 20250181576 A1).
Rejection under 35 U.S.C. 103
A(i) The combination of Elnaz and Eghbali does not arrive at the amended claims
Applicant argues that “Elnaz does not teach the amended independent claim limitations of generating a prompt with a textual tool listing, receiving an LLM response indicating tools from the listing, executing those tools to obtain context, generating a second prompt with accumulated context, and determining by a host program whether to iterate or output a final response. Eghbali does not cure these deficiencies.”. The examiner agrees. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground of rejection has been given over Enlaz and further in view of Eghbali, Keller and Tyagi.
A(ii) Claims 2/9/16: Generating computer code in an application development environment
Applicant argues that “A reference that teaches code generation in an IDE does not teach the specific tool-listing, LLM-directed tool selection, and iterative context provisioning architecture of the independent claims.”. The examiner agrees that Eghbali does not teach the specific tool-listing, context provisioning architecture. However, the rejection does not rely upon Eghbali for the tool-listing, iterative context provisioning limitations. These are taught by the combination of Enlaz, Keller, and Tyagi. Eghbali is relied upon only for teaching generating computer code in an application development environment.
A(iii) Claims 3/10/17: Listing comprises name, description, and example syntax
Applicant argues that “The USPTO relied on Eghbali Definition 3.1 for the limitation that the listing of software tools comprises, for each software tool, a name, a description, and example syntax. Applicant respectfully submits that Eghbali does not teach this limitation.”. Examiner does not agree. Eghbali teaches function references that include a name, available type annotations, and function-level documentation (3.1, 3.2.1). A person of ordinary skill in the art would have understood the function name, as corresponding to the claimed name and the function level documentation as corresponding to the claimed description. Furthermore, the disclosed function signatures and parameter information provide example usage information for invoking the functions. Accordingly, Eghbali teaches a listing comprising a name, description, and example syntax associated with each function. Therefore, the cited combination teaches the claimed limitation.
A(iv) Claims 6/13/20: Iterative Prompting with finality determination
Applicant argues that Eghbali does not teach the limitation that “the second response comprises a second indication of a second subset of tools, the second subset is executed for second context, a third prompt is generated, and the third response is utilized as a final response based on a determination that it contains an indication that it is final.”. The examiner agrees that Eghbali does not teach this limitation. However, Tyagi teaches determining whether additional executions are required based upon generated results, where the results from previous iterations carry over to future iterations, and further teaches outputting a final result when no additional executions are needed (0047, 0052-0054, 0064). Accordingly, the cited combination teaches determining whether to continue iterating or provide a final response.
A(v). No motivation to combine
Applicant argues “Even if Elnaz and Eghbali individually taught the claimed limitations, which they do not, a person of ordinary skill in the art would not have been motivated to combine them.”. The previous rejection over Elnaz and Eghbali has been withdrawn. However, claims 1, 5, 12, and 19 are rejected under 35 U.S.C. 103 over Enlaz in view of Eghbali, Keller, and Tyagi for the reasons set forth herein. It would have been obvious to a person of ordinary skill in the art before the effective filing data to modify Enlaz in view of Eghbali, Keller, and Tyagi because doing so would have predictably improved the ability of the system to select appropriate software tools, obtain additional context through iterative processing, and generate a more complete response to request. Each references teaches known techniques directed to the same general problem of improving LLM-assisted task execution. Combining these known techniques according to their established functions would have yielded predictable results. See KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398 (2007).
B(i). Patentability of base claims
Applicant argues that “Claims 4, 11, and 18 are therefore patentable at least for the reasons set forth with respect to their base claims.”. Examiner does not agree. As set forth above the rejection of the base claims is maintained. Claims 4, 11, and 18 depend from these claims and therefore are not patentable for the reasons set forth with respect to their base claims.
B(ii). Albert does not teach the claimed limitation
Applicant argues that “Albert does not describe this component as a software tool that is part of a textual listing provided to an LLM in a prompt. Albert does not describe this component as something that an LLM can select from a listing. Albert does not describe this component as something that a host program executes in response to an LLM's selection to gather context for an iterative prompting loop”. Examiner does not agree that this combination fails to arrive at the claimed limitation. Bates (Mistakenly identified as Albert by the Examiner) teaches displaying the current cursor location within code being edited (0057). Claim 4 does not require the reference teaching the code-location function to disclose the entire textual-listing architecture in the parent claim. Rather, the rejection relies on Enlaz, Keller, and Tyagi for the aforementioned architecture, while Bates is relied on for the limitation of returning the current location of computer code. Therefore, the cited combination teaches the claimed limitation.
B(iii). No Motivation to combine three disparate references
Applicant argues that “A person of ordinary skill working on LLM-based context provisioning would not have been motivated to extract a code location identification feature from Albert's development environment and reconstitute it as a callable software tool within a textual listing provided to an LLM for iterative context provisioning.”. Examiner does not agree. A person of ordinary skill in the art would have been motivated to incorporate such code-location information into the combined Enlaz, Keller, and Tyagi system in order to provide additional context relevant to software-development tasks, improving the relevance and accuracy of responses.
C(i) Patentability of the base claims
Applicant argues that “Claims 7 and 14 are therefore patentable at least for the reasons set forth with respect to their base claims.”. Examiner does not agree. As set forth above the rejection of the base claims is maintained. Claims 7 and 14 depend from these claims and therefore are not patentable for the reasons set forth with respect to their base claims.
C(ii). Taihong does not teach the claimed limitations in the context of the claimed architecture
Applicant argues that “Taihong Does Not Teach the Claimed Limitations in the Context of the Claimed Architecture” because “Taihong does not teach the specific iterative architecture of claims 6/13”. Examiner agrees that Taihong does not teach the specific iterative architecture. Taihong is not relied upon for the iterative tool-selection and context-provisioning limitations but instead for teaching a code-development platform operating within an application development environment and transmitting requests to, and receiving responses from an LLM. The iterative tool-selection and context-provisioning limitations are taught by the combination of Enlaz, Keller, and Tyagi. Therefore, the cited combination teaches the claimed limitation.
C(iii) No motivation to combine three references
The applicant argues that “The USPTO's stated motivation, "to improve the accuracy of code generation" (Office Action at p. 14), is generic and does not explain why a person of ordinary skill would specifically combine these three references or how doing so would arrive at the claimed iterative tool-selection architecture operating within an ADE.”. Examiner does not agree. Taihong is relied upon for teaching operation within an Application Development Environment and transmitting requests to, and receiving responses from, a large language model. A person of ordinary skill in the art would have been motivated to incorporate the teachings of Taihong in to the combined Enlaz, Keller, Tyagi system in order to apply the iterative tool-selection system in a software-development setting, improving code generation and providing contextual information during development.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 5, 12, 15, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Enlaz (WO 2025141192 / US Provisional: 63/616,230) in view of Keller (US 20250156535 A1) and Tyagi (US 20250181576 A1).
Regarding claim 1, Enlaz teaches a system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform operations comprising (Fig. 2, teaches the whole system): receiving a request to perform a computer task, the request comprising an identification of a first software tool (Para 0043, “one or more non-transitory computer readable media storing embeddings generated by performing operations including maintaining software tool use data that includes a software tool selection embedding and a respective software tool embedding for each software tool in a set including multiple software tools.”), executing the subset in the first response to obtain context regarding the request (Para 0025, “obtaining an output from the software tool” where the output is the result of executing the context gathering tool).
Enlaz does not teach generating a first prompt comprising a set of instructions, a textual listing of a plurality of software tools, and the request; sending the first prompt to a large language model (LLM), receiving a first response from the LLM, the first response comprising an indication of a subset of the plurality of software tools in the listing.
However, Keller teaches generating a first prompt comprising a set of instructions (system prompt, Para 0011), a textual listing of a plurality of software tools (available tools, Para 0011), and the request (Para 0011, “The application provides the user prompt, a system prompt, and a list of available tools to an LLM for guidance on how to accomplish the target action.” Wherein the request is the user prompt); sending the first prompt to a large language model (LLM) (Para 0011, “The application provides the user prompt … to an LLM “), receiving a first response from the LLM, the first response comprising an indication of a subset of the plurality of software tools in the listing. (Para 0011, “In response, the LLM generates a first output directing the application to use a web tool from the list of tools”, wherein one web tool comprises a subset of a larger set of tools).
It would have been obvious to a person of ordinary skill in the art before the effective filing date to modify Enlaz such as to incorporate the teachings of Keller in order to allow the LLM to select from a predetermined set of tools in a convenient single prompt. (Para 0011).
Enlaz modified by Keller teaches a prompt comprising the set of instructions, the listing, and the request (Para 0011). Sending a prompt to the LLM, and receiving a response from the LLM (Para 0011, “In response, the LLM generates a first output directing the application to use a web tool from the list of tools”).
Enlaz modified by Keller does not teach wherein the system iterates a second time, and wherein the prompt includes the context from previous iterations; and determining, by the host program, based on the second response, whether to iterate by generating an additional prompt or to output the second response as a final response.
However, Tyagi teaches wherein the system iterates a second time (Para 0047, “the task agent can either output an indication that no more executions of applications are needed or an identification of a second application or a second series of applications to use or execute to satisfy the one or more tasks.”), and wherein the prompt includes the context from previous iterations (Para 0027, “The task agent can iteratively repeat this process to generate new results until the task agent determines an aggregation of all of the generated results satisfy the one or more tasks that correspond to the request for information.”); determining, by the host program, (the analytics server) based on the second response, whether to iterate by generating an additional prompt to output the second response as a final response. (Para 0009, “the task agent iteratively determines whether to execute another application based on whether all generated results satisfy the one or more tasks” and para 0047, “Based on the execution, the task agent can either output an indication that no more executions of applications are needed or an identification of a second application or a second series of applications to use or execute to satisfy the one or more tasks.” Where the task agent is directed by the analytics server as in para 0052).
It would have been obvious to a person of ordinary skill in the art before the effective filing date to modify Enlaz such as to incorporate the teachings of Tyagi as doing so improves task completion by obtaining sufficient context before generating a final result. (Para 0009).
Regarding claim 5, Enlaz teaches the system wherein the indication of a subset comprises
an indication of the first software tool (Para 0024: “a software tool selection output that identifies
a particular software tool from the set of software tools”);
Claim 8 & 15 are analogous to claim 1 in that they recite substantially the same limitations. They are therefore rejected for the same reasons set forth above.
Claim 12 & 19 are analogous to claim 5 in that they recite substantially the same limitations. They are therefore rejected for the same reasons set forth above.
Claims 2-3, 9-10, 16-17 are rejected under 35 U.S.C. 103 as being unpatentable over Enlaz (WO 2025141192 / US Provisional: 63/616,230), Keller (US 20250156535 A1) and Tyagi (US 20250181576 A1) as above in claims 1, 5, 8, 12, 15 and 19 and further in view of Eghbali et al.
Regarding claim 2, Enlaz does not teach the system wherein the computer task comprises
generating computer code in an application development environment.
However, Eghbali does teach the system wherein the computer task comprises generating
computer code in an application development environment. (Fig 3: the output is generated
computer code, 3.2.4: “an IDE-based implementation”, where an integrated development
environment (IDE) can be interpreted as an application development environment).
It would have been obvious to a person of ordinary skill in the art to modify Enlaz before
the effective filing date in such a way as to incorporate the teachings of Eghbali to improve development efficiency by enabling generation of code in an application development environment.
Regarding claim 3, Enlaz does not teach the system wherein the listing of the plurality of
software tools comprises, for each software tool, a name, a description, and example syntax.
However, Eghbali teaches the system wherein the listing of the plurality of software tools
comprises, for each software tool, a name, a description, and example syntax. (Fig 4., Definition
3.1: function reference includes name, type annotations, available docstring).
It would have been obvious to a person of ordinary skill in the art before the effective
filing date to modify Enlaz in such a way as to incorporate the teachings of Eghbali in order to provide additional information regarding functions and their usage, improving a user’s ability to generate code. (3.1).
Claim 9 & 16 are analogous to claim 2 in that they recite substantially the same limitations. They are therefore rejected for the same reasons set forth above.
Claim 10 & 17 are analogous to claim 3 in that they recite substantially the same limitations. They are therefore rejected for the same reasons set forth above.
Claims 4, 11, 18 are rejected under 35 U.S.C. 103 as being unpatentable over Enlaz (WO 2025141192 / US Provisional: 63/616,230), Keller (US 20250156535 A1) and Tyagi (US 20250181576 A1) as above in claims 1, 5, 6, 8, 12, 13, 15, 19, and 20 and further in view of Bates et al. (US20080263517).
Regarding claim 4, Enlaz modified by Keller and Tyagi teaches a plurality of software tools in the listing.
However, Bates teaches a function that returns a current location in computer code that is being edited by a user with the application development environment. (Para 0057, “The source editor then displays the current source, which is typically only a portion of the source at the current location of the cursor”).
It would have been obvious to a person of ordinary skill in the art before the effective
filing date to modify Enlaz in such a way as to incorporate the teachings of Bates in order to improve provide information regarding current editing location, thereby improving the relevance and accuracy of generated code (Para 0057).
Claim 11 & 18 are analogous to claim 4 in that they recite substantially the same limitations. They are therefore rejected for the same reasons set forth above.
Claims 7 & 14 are rejected under 35 U.S.C. 103 as being unpatentable over Enlaz (WO 2025141192 / US Provisional: 63/616,230), Keller (US 20250156535 A1), Tyagi (US 20250181576 A1) as above in claims 1, 5, 6, 8, 12, 13, 15, 19, and 20 and further in view of Taihong et al. (WO2025145584).
Regarding claim 7, Enlaz modified by Keller and Tyagi does not teach the system wherein the request is received from an Application Development Environment and wherein the utilizing comprises sending the third response to the application development environment.
However, Taihong teaches the system wherein the request is received from an
Application Development Environment and wherein the utilizing comprises sending the third
response to the application development environment. (Summary of invention Para 2, “The
method can be executed by a code development platform. The code development platform
supports code generation.”).
It would be obvious to a person of ordinary skill in the art before the effective filing date
to modify Enlaz in such a way as to incorporate the teachings of Taihong in order to improve integration of code-generation functionality into a software development workflow (Summary).
Claim 14 is analogous to claim 7 in that it recites substantially the same limitations. It is therefore rejected for the same reasons set forth above.
Allowable Subject Matter
Claims 6, 13 and 20 are being objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MICHAEL ALAN FOSTER JR. whose telephone number is (571)272-8874. The examiner can normally be reached M - F 8:00am - 6:00pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Hai Phan can be reached at (571) 272-6338. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MICHAEL A FOSTER JR/Examiner, Art Unit 2654
/HAI PHAN/Supervisory Patent Examiner, Art Unit 2654
Prosecution Insights