{"id":50112329,"url":"https://github.com/zhutao100/macos-private-framework-research-skill","last_synced_at":"2026-05-23T13:04:20.561Z","repository":{"id":359181148,"uuid":"1244915957","full_name":"zhutao100/macos-private-framework-research-skill","owner":"zhutao100","description":"Agents skill to research modern macOS private frameworks with dyld shared cache extraction, Objective-C/Swift metadata reconstruction, Hopper/RuntimeViewer/RuntimeBrowser/ipsw workflows, and MOTIF-style LLM-guided type inference","archived":false,"fork":false,"pushed_at":"2026-05-20T23:13:56.000Z","size":169,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-05-20T23:55:59.522Z","etag":null,"topics":["agents-skills","macos","macos-private-frameworks","reverse-engineering"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/zhutao100.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null}},"created_at":"2026-05-20T18:16:48.000Z","updated_at":"2026-05-20T23:14:00.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/zhutao100/macos-private-framework-research-skill","commit_stats":null,"previous_names":["zhutao100/macos-private-framework-research-skill"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/zhutao100/macos-private-framework-research-skill","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhutao100%2Fmacos-private-framework-research-skill","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhutao100%2Fmacos-private-framework-research-skill/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhutao100%2Fmacos-private-framework-research-skill/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhutao100%2Fmacos-private-framework-research-skill/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/zhutao100","download_url":"https://codeload.github.com/zhutao100/macos-private-framework-research-skill/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zhutao100%2Fmacos-private-framework-research-skill/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33396584,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-23T04:15:53.637Z","status":"ssl_error","status_checked_at":"2026-05-23T04:15:53.242Z","response_time":53,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agents-skills","macos","macos-private-frameworks","reverse-engineering"],"created_at":"2026-05-23T13:04:19.732Z","updated_at":"2026-05-23T13:04:20.556Z","avatar_url":"https://github.com/zhutao100.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# macOS Private Framework Research Skill\n\nInstallable agent skill repository for researching modern macOS private frameworks with a repeatable, evidence-backed workflow.\n\n## Skill\n\n`macos-private-framework-research` helps agents discover private-framework dependencies, extract dyld shared cache images, reconstruct Objective-C/Swift interfaces, triage underspecified signatures, and run an LLM-assisted type-inference loop with static validation.\n\nPrimary entrypoint:\n\n```text\nmacos-private-framework-research/SKILL.md\n```\n\n## Clone\n\nReference checkouts are optional source signposts. Avoid initializing them from `git clone --recurse-submodules --shallow-submodules`: Git runs recursive submodule checkout with `--no-single-branch`, which makes high-ref repositories fetch every branch/tag tip at depth 1.\n\nUse a two-step checkout when reference sources are needed:\n\n```bash\ngit clone -o zhutao100 https://github.com/zhutao100/macos-private-framework-research-skill.git\ncd macos-private-framework-research-skill\ngit submodule update --init --recursive --depth 1 --single-branch --jobs 4\n```\n\n`reference-checkout/hammerspoon` is opt-in because the recursive clone path is especially large. Initialize it only when needed:\n\n```bash\ngit submodule update --init --depth 1 --single-branch --checkout -- reference-checkout/hammerspoon\n```\n\n## Install\n\nUser-scoped install for Codex CLI and other Open Agent Skills-compatible clients:\n\n```bash\nmacos-private-framework-research/scripts/install_codex_skill.sh --replace\n```\n\nRepository-scoped install:\n\n```bash\nmacos-private-framework-research/scripts/install_codex_skill.sh --scope repo --replace\n```\n\nManual install:\n\n```bash\nmkdir -p \"$HOME/.agents/skills\"\ncp -R macos-private-framework-research \"$HOME/.agents/skills/\"\n```\n\n## Common Commands\n\nInventory host, dyld cache, and tools:\n\n```bash\nmacos-private-framework-research/scripts/macos_private_framework_inventory.py \\\n  --output /tmp/macos-pf-inventory.md \\\n  --json-output /tmp/macos-pf-inventory.json\n```\n\nResolve optional non-built-in toolchains only when a workflow needs them:\n\n```bash\nmacos-private-framework-research/scripts/resolve_toolchains.py ipsw dyld-shared-cache-extractor\n```\n\nDiscover private-framework usage by a system app or client binary:\n\n```bash\nmacos-private-framework-research/scripts/discover_private_frameworks.py \\\n  --output /tmp/disk-utility-private-frameworks.md \\\n  --json-output /tmp/disk-utility-private-frameworks.json \\\n  \"/System/Applications/Utilities/Disk Utility.app\"\n```\n\nExtract a framework from the dyld shared cache:\n\n```bash\nmacos-private-framework-research/scripts/extract_dyld_framework.sh \\\n  --framework DiskManagement \\\n  --output-dir /tmp/macos-private-frameworks\n```\n\nUse `--enrich-objc-stubs` only when extraction-time enrichment is needed; run `ipsw class-dump` separately for bounded header output.\n\nBuild a capped framework manifest before loading raw dependency, symbol, or string dumps:\n\n```bash\nmacos-private-framework-research/scripts/framework_macho_manifest.py \\\n  --framework DiskManagement \\\n  --json-output /tmp/DiskManagement.manifest.json \\\n  --markdown-output /tmp/DiskManagement.manifest.md \\\n  --cache-evidence\n```\n\nCollect focused entitlement evidence before loading raw plist dumps:\n\n```bash\nmacos-private-framework-research/scripts/collect_code_entitlements.py \\\n  --focus-pattern 'diskmanagement|mach-lookup|xpc|sandbox' \\\n  --output /tmp/DiskManagement.entitlements.md \\\n  --json-output /tmp/DiskManagement.entitlements.json \\\n  /path/to/candidate-binary\n```\n\nProbe candidate C symbols and Objective-C classes without calling them:\n\n```bash\nmacos-private-framework-research/scripts/dlopen_symbol_probe.swift \\\n  --image /System/Library/PrivateFrameworks/DiskManagement.framework/DiskManagement \\\n  --symbol CandidateFunction \\\n  --class CandidateClass \\\n  --json\n```\n\nTriage reconstructed headers for underspecified method signatures:\n\n```bash\nmacos-private-framework-research/scripts/objc_header_triage.py \\\n  --headers /tmp/DiskManagement.headers \\\n  --output /tmp/DiskManagement.candidates.md \\\n  --json-output /tmp/DiskManagement.candidates.json\n```\n\nMarkdown reports are bounded by default for agent ingestion; JSON outputs remain complete.\n\nBuild an LLM-ready MOTIF-style inference context for one candidate:\n\n```bash\nmacos-private-framework-research/scripts/build_motif_context.py \\\n  --candidate-json /tmp/DiskManagement.candidates.json \\\n  --candidate-id 1 \\\n  --headers /tmp/DiskManagement.headers \\\n  --binary /tmp/macos-private-frameworks/System/Library/PrivateFrameworks/DiskManagement.framework/Versions/A/DiskManagement \\\n  --output /tmp/DiskManagement.candidate-1.context.json \\\n  --prompt-output /tmp/DiskManagement.candidate-1.prompt.md\n```\n\nLint candidate reconstructed headers before treating them as evidence:\n\n```bash\nmacos-private-framework-research/scripts/objc_signature_linter.py \\\n  --headers /tmp/DiskManagement.inferred.headers \\\n  --output /tmp/DiskManagement.lint.md \\\n  --json-output /tmp/DiskManagement.lint.json \\\n  --compile\n```\n\n## Validation\n\nRepository validation:\n\n```bash\nmacos-private-framework-research/scripts/validate_skill_repo.py .\nmacos-private-framework-research/scripts/resolve_toolchains.py --json-output /tmp/macos-pf-toolchains.json \\\n  \u003e/tmp/macos-pf-toolchains.md\nmacos-private-framework-research/scripts/framework_macho_manifest.py --framework IntelligenceFlow --cache-evidence \\\n  --json-output /tmp/macos-pf-framework-manifest.json \\\n  --markdown-output /tmp/macos-pf-framework-manifest.md\nmacos-private-framework-research/scripts/collect_code_entitlements.py --focus-pattern 'intelligenceflow|biome|mach-lookup' \\\n  --output /tmp/macos-pf-entitlements.md \\\n  --json-output /tmp/macos-pf-entitlements.json \\\n  /System/Library/PrivateFrameworks/IntelligenceFlowRuntime.framework/Versions/A/intelligenceflowd\npython3 -m py_compile macos-private-framework-research/scripts/*.py\nbash -n macos-private-framework-research/scripts/*.sh\nmacos-private-framework-research/scripts/dlopen_symbol_probe.swift --help \u003e/tmp/macos-pf-dlopen-help.txt\nzsh -n framework-surveys/intelligenceflow-agent-survey/scripts/*.zsh\npython3 -m py_compile framework-surveys/intelligenceflow-agent-survey/scripts/*.py\nzsh -n framework-surveys/skylight-agent-survey/tools/*.zsh\npython3 -m py_compile framework-surveys/skylight-agent-survey/tools/*.py\n```\n\nOn a macOS host, also run the inventory script, a small app/binary discovery pass, and survey header verifiers when touching the survey packages.\n\n```bash\nframework-surveys/intelligenceflow-agent-survey/scripts/verify_intelligenceflow_presence_header.zsh \\\n  /tmp/intelligenceflow-presence-verify.json\nswift framework-surveys/skylight-agent-survey/tools/dlopen_probe_symbols.swift --json \\\n  \u003e/tmp/skylight-dlsym-probe.json\nframework-surveys/skylight-agent-survey/tools/verify_skylight_readonly_header.zsh \\\n  /tmp/skylight-readonly-verify.json\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fzhutao100%2Fmacos-private-framework-research-skill","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fzhutao100%2Fmacos-private-framework-research-skill","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fzhutao100%2Fmacos-private-framework-research-skill/lists"}