You are helping the user reduce the size of AI instruction files — skills, CLAUDE.md, slash commands, agent definitions, and other markdown prompts — by removing the least important content while preserving meaning and intent.
/shrink <percentage> - percentage to reduce byApply these techniques in order of priority to reach the target reduction:
| # An example for a config file which a developer might while work. | |
| # It allows the developer to have 1 user in 1 AWWS Account, | |
| # and access multiple other accounts (which exist for different environments). | |
| # The access to different environment is being done with AssumeRole protected with MFA. | |
| # The `credential_process` allows simple usage of this config file by `aws --profile john` | |
| # or IDEs, and in this example the session tokens for each profile will be returned from | |
| # aws-vault which stores the long-lived credentials in a secure keychain. | |
| [default] | |
| region=us-east-1 |
| """A python example which shows the time complexity of the counting sort algorithm (O(n + k)). | |
| The final output is a table showing the execution time of the counting sort algorithm for lists of different lengths. | |
| Copied from my execution on my machine: | |
| 524288 (2**19) -> 0.25 | |
| 1048576 (2**20) -> 0.49 | |
| 2097152 (2**21) -> 0.92 | |
| 4194304 (2**22) -> 1.75 | |
| 8388608 (2**23) -> 3.47 |
| """As part of the 'Data Structures and Introduction to Algorithms' course, I was asked to solve the following problem: | |
| Write a function that sorts five elements in seven comparisons in the worst case. | |
| As I found this problem interesting, I decided to implement it in Python. For better understanding, I recommend | |
| reading the answer from the link below, which explains the algorithm. | |
| """ | |
| import random | |
| from typing import List |
| name: Secrets trial | |
| on: | |
| workflow_dispatch: | |
| inputs: | |
| secrets_list: | |
| required: true | |
| description: 'List of secrets to pass to action' | |
| jobs: | |
| tmp-secrets: |
| import datetime | |
| import os | |
| import requests | |
| from github import Github | |
| REPO_NAME = 'some-repo' | |
| REPO_OWNER = 'some-owner' | |
| LAST_WF_DATE_TO_SHOW = datetime.datetime(2023, 7, 2, 9, 11) |
| # A configured profile can already exist or not. If not, aws-vault will create the profile in `~/.aws/config`. | |
| > aws-vault add some-developer | |
| Enter Access Key Id: *** | |
| Enter Secret Access Key: *** |
| > aws sts get-session-token | |
| { | |
| "Credentials": { | |
| "AccessKeyId": "***", | |
| "SecretAccessKey": "***", | |
| "SessionToken": "***", | |
| ... | |
| } | |
| } |
| > aws-vault exec dev -- aws lambda invoke --function-name some-function --payload '{}' | |
| > aws-vault exec staging -- aws lambda list-functions |
| [default] | |
| region = us-east-1 | |
| mfa_serial = arn:aws:iam::1000:mfa/some-developer | |
| [profile some-developer] | |
| Credential_process = aws-vault export --format=json some-developer | |
| [profile dev] | |
| role_arn = arn:aws:iam::2000:role/Admin | |
| source_profile = some-developer |