TeamPCP has expanded a multi-ecosystem software supply chain campaign that compromised open-source security and developer tools including Trivy, Checkmarx KICS, LiteLLM, Telnyx, GitHub Actions, OpenVSX extensions, Docker images, and packages published through PyPI and npm. Reporting indicates the attackers used stolen developer and publishing credentials to push malicious releases through trusted channels, harvest environment variables, shell histories, cloud credentials, and GitHub tokens, and move laterally across CI/CD environments. In the Telnyx incident, valid credentials were reportedly used to publish malicious PyPI releases, with a second-stage payload hidden in a WAV file and code triggered on import.
The campaign is now being linked to follow-on ransomware activity through an alleged partnership between TeamPCP and the Vect ransomware group, which has been advertised on BreachForums as an emerging ransomware-as-a-service operation. Researchers say the supply chain compromises may serve as initial access for extortion campaigns against downstream organizations, with TeamPCP reportedly recruiting negotiators after the Trivy breach and previously exfiltrating roughly 300 GB of compressed credentials; the LiteLLM compromise alone was tied to hundreds of thousands of stolen credentials. The incidents underscore how compromised open-source tooling and CI/CD infrastructure can give attackers privileged enterprise access and create a path from package poisoning to ransomware deployment.

Mallory correlates global threat intelligence with your attack surface — know if you’re exposed before adversaries strike.
62 events from the most recent confirmed update back to the earliest known activity.
A June 11, 2026 threat research report described a supply-chain attack involving malicious DurableTask packages on PyPI that deployed multi-stage malware targeting developer and cloud-native environments. The analysis said the payload added persistence, stole credentials across multiple cloud platforms, abused GitHub tokens, targeted Kubernetes and HashiCorp Vault, used resilient command-and-control discovery, and included geo-targeted destructive capabilities.
On 2026-06-08, a new Shai-Hulud campaign wave reportedly compromised 37 Python wheels across 19 PyPI packages. Zscaler said the malicious packages embedded prompt injection intended to mislead LLM-based security scanners, while still abusing trusted publishing and valid provenance from trusted environments.
On 2026-06-05, a malicious commit was pushed to Microsoft's Azure/durabletask GitHub repository using a previously compromised contributor account, adding Claude Code, Gemini CLI, Cursor, and VS Code configuration files that executed embedded JavaScript to harvest credentials when a developer opened the repository. GitHub responded by disabling 73 repositories across four Microsoft GitHub organizations, and the report linked the incident to the broader Miasma infrastructure associated with TeamPCP.
Beginning on 2026-06-01, attackers launched a major npm supply-chain wave using TeamPCP-linked tradecraft, including the Miasma compromise of dozens of @redhat-cloud-services packages. The same wave reportedly included the Phantom Gyp follow-on campaign affecting 57 more packages through binding.gyp and node-gyp install-time execution, with attribution described as less certain due to copycat abuse of the leaked Mini Shai-Hulud framework.
On 2026-05-28, CISA reportedly issued a standalone advisory covering the Nx Console compromise, GitHub repository theft, and related Megalodon GitHub Actions abuse after adding key campaign vulnerabilities to its Known Exploited Vulnerabilities catalog on 2026-05-27. This marked the first formal standalone U.S. government advisory action described for the TeamPCP supply-chain campaign.
A SlowMist report published on 2026-05-22 said the May 2026 Shai-Hulud/Mini Shai-Hulud attack chain began with a TanStack CI trust-boundary failure involving a pull_request_target workflow, GitHub Actions cache poisoning, and extraction of OIDC tokens from runner memory. The report said this enabled malicious TanStack package releases and subsequent credential theft and propagation across additional ecosystems.
On 2026-05-21, GitHub disclosed that TeamPCP gained access to at least 3,800 repositories after a developer installed a poisoned VS Code extension. GitHub said the affected repositories contained GitHub’s own code rather than customer code, adding a major newly disclosed victim and a concrete scale estimate for the campaign’s repository compromise.
On 2026-05-21, Hunt.io reported on a 13-file Python toolkit used by TeamPCP after a supply-chain compromise. The analysis said the toolkit could continue operating even after command-and-control takedowns by relying on FIRESCALE, GitHub, and victims' own accounts for exfiltration and operational continuity, adding new infrastructure and tradecraft details not covered in earlier reporting.
On 2026-05-15, Akamai reported that a new Mini Shai-Hulud variant had escalated TeamPCP’s credential-theft and propagation activity with greater automation and harvesting across repositories, packages, and developer tooling. The report said the worm and supporting toolchain were now public, raising the risk that other threat actors could reuse the same supply-chain attack techniques.
On 2026-05-14, reporting said TeamPCP and BreachForums launched a contest offering a $1,000 Monero prize for compromising open-source packages with the Shai-Hulud attack tool. The contest allegedly ranks participants by infected packages' download counts, marking a new crowdsourcing and recruitment effort aimed at expanding copycat software supply-chain attacks.
On 2026-05-13, Trend Micro published new technical analysis of TeamPCP activity involving the elementary-data Python package, describing host reconnaissance, credential and secret theft, AWS Secrets Manager and SSM abuse, staging into trin.tar.gz, and HTTPS exfiltration using a custom header. The report also said TeamPCP was increasingly using lower-friction initial access such as GitHub workflow command injection through unsanitized user-controlled expressions, and provided new hunting and mitigation guidance.
On 2026-05-11, a coordinated Mini Shai-Hulud supply-chain attack attributed to TeamPCP reportedly compromised more than 170 npm and PyPI packages, including packages tied to TanStack, Mistral AI, and OpenSearch. Expel said the malware abused GitHub Actions pull_request_target behavior and OIDC token extraction to mint valid publish tokens and ship malicious updates with apparently valid SLSA Build Level 3 provenance attestations, while stealing credentials and modifying VS Code and Claude Code settings for persistence.
On 2026-05-11, Checkmarx confirmed that a malicious modified version of its Jenkins AST plugin had been published to the Jenkins Marketplace. The company advised users to use version 2.0.13-829.vc72453fa_1c16 from December 17, 2025 or earlier and released clean version 2.0.13-848.v76e89de8a_053 on GitHub and the Jenkins Marketplace.
On 2026-05-09, Wiz listed an incident involving the Checkmarx Jenkins AST Plugin and attributed the compromise to TeamPCP. The supplied reference provides no further technical, impact, or remediation details beyond identifying the affected plugin and actor.
A May 2026 report alleged that Mercor's breach tied to the March 24 LiteLLM compromise resulted in the theft of about 4 TB of data. The claimed haul included source code, user data, contractor information, and AI-related configuration artifacts such as MCP files, marking a significant escalation in the reported impact on Mercor.
Between 2026-04-29 and 2026-04-30, a self-propagating software supply-chain campaign dubbed Mini Shai-Hulud reportedly began with four official SAP npm packages and spread into PyTorch Lightning and Intercom-related packages across npm, PyPI, and Packagist. Reporting attributed the operation to TeamPCP with high confidence and said it stole developer and cloud credentials, targeted AI coding agent configuration files for persistence, and was linked to about 1,800 GitHub repositories created using stolen credentials.
On 2026-04-27, Checkmarx disclosed that data posted online after LAPSUS$ leak claims appeared to have originated from one of its GitHub repositories. The company said the access was likely tied to the broader TeamPCP-linked supply-chain activity that had already compromised its KICS-related tooling.
By 2026-04-27, reporting identified an npm worm cluster dubbed CanisterSprawl associated with the late-April TeamPCP-linked supply-chain activity. The update added new technical detail about malware used in the Checkmarx KICS, Bitwarden CLI cascade, and xinference-related compromise wave.
On 2026-04-24, Sophos X-Ops reported that the April 22 Checkmarx KICS and Bitwarden CLI supply-chain compromises were part of a coordinated campaign using the same command-and-control domain, audit.checkmarx[.]cx (94.154.172[.]43). The analysis added new technical detail that the Bitwarden payload abused stolen GitHub tokens to inject malicious workflows and create public repositories as dead drops, while both payloads targeted developer credentials, cloud secrets, and AI assistant configuration files.
On 2026-04-22, Bitwarden confirmed that attackers compromised its CI/CD pipeline and briefly published a malicious @bitwarden/cli@2026.4.0 package to npm. The company said the package was available between 5:57 PM and 7:30 PM ET, revoked compromised access, deprecated the release, began remediation, and reported no evidence that end-user vault data, production data, or production systems were affected.
On 2026-04-21, attackers reportedly published malicious xinference releases to PyPI during the same wave that included the Checkmarx KICS Docker Hub compromise and Bitwarden CLI cascade. The report said TeamPCP denied involvement despite strong code and tradecraft similarities, leaving attribution unresolved between TeamPCP, a copycat, or a false-flag operation.
Between 2026-04-21 and 2026-04-22, reporting said TeamPCP's 26-day pause in active compromises ended when attackers used valid publisher credentials to compromise Checkmarx's official KICS Docker Hub repository. The malicious images reportedly also seeded trojanized VS Code/Open VSX extensions and contributed to the downstream Bitwarden CLI npm compromise through poisoned CI/CD dependencies.
On 2026-04-15, Vect reportedly published its first victim on its leak site, identifying a property-management technology company and claiming to have stolen about four million emails and 700 GB of data. The post marked a concrete shift from earlier TeamPCP-Vect partnership claims into active double-extortion operations tied to the Trivy supply-chain campaign.
On 2026-04-08, the CISA Known Exploited Vulnerabilities remediation deadline for CVE-2026-33634 arrived. Reporting said no standalone U.S. government advisory specific to TeamPCP had been issued by that date.
By 2026-04-08, Google Threat Intelligence Group had formally designated TeamPCP as UNC6780 and identified its credential stealer as SANDCLOCK. This added a new attribution label and malware naming detail for the actor behind the Trivy, Checkmarx, LiteLLM, and Telnyx compromises.
By 2026-04-08, Elastic Security Labs had published new detection guidance focused on TeamPCP's container-related attack activity. The update added fresh defensive and technical detail for identifying the group's ongoing post-compromise behavior.
By 2026-04-08, reporting said the Trivy-linked intrusion at Cisco had escalated beyond initial access, with attackers reportedly cloning more than 300 private repositories, stealing AWS keys, and taking source code from build systems and developer workstations. The update also noted unverified extortion claims tied to the stolen Cisco data.
On 2026-04-03, reporting said Sportradar AG suffered a systemic compromise through the Trivy supply-chain vector. VECERT attributed the operation jointly to TeamPCP and Vect ransomware and said exposed data included personal information, client records, and production credentials.
On 2026-04-03, CERT-EU said the European Commission’s Trivy-linked Europa.eu AWS breach involved about 92 GB of compressed stolen data and could affect at least 29 other EU entities as well as dozens of internal Commission clients. The agency said nearly 52,000 files contained sent email messages, raising potential personal-data exposure risks.
On 2026-04-03, CERT-EU confirmed that the European Commission’s Europa web hosting platform on AWS was breached as a result of the Trivy supply-chain compromise tracked as CVE-2026-33634. This added the European Commission as a newly disclosed government victim in the broader TeamPCP-linked campaign.
By 2026-04-03, reporting said Mercor AI was facing a class action investigation following its disclosure that it had been breached through the TeamPCP-linked LiteLLM supply-chain compromise. This marked a new legal and business-impact development beyond the earlier breach confirmation and scope estimates.
On 2026-04-02, Mercor said it was one of thousands of companies impacted by the LiteLLM supply-chain attack and that it had contained and remediated the incident while its forensic investigation continued. The same reporting cited researchers saying more than 1,000 SaaS environments were affected, marking a major escalation in the known downstream scope of the TeamPCP-linked campaign.
On 2026-04-01, reporting said the axios npm compromise was attributed to North Korean actor UNC1069 rather than TeamPCP. Analysts noted the stolen npm token may still have originated from the broader credential ecosystem seeded by TeamPCP, but the compromise itself was no longer assessed as a TeamPCP operation.
On 2026-04-01, ownCloud disclosed that its build infrastructure was affected by the Trivy supply-chain compromise tracked as CVE-2026-33634. This added another publicly identified downstream victim tied to the broader TeamPCP-linked campaign.
On 2026-04-01, BerriAI said LiteLLM publishing had resumed after a Mandiant-led forensic audit found that only versions 1.82.7 and 1.82.8 were malicious. The statement marked a concrete remediation and recovery step following TeamPCP's March 24 compromise of LiteLLM.
On 2026-04-01, incident response findings described TeamPCP validating stolen credentials with TruffleHog and then rapidly enumerating AWS and Azure environments. The activity focused on IAM, compute, storage, databases, and container infrastructure, adding new technical detail about the group’s post-compromise tradecraft.
On 2026-04-01, Mercor AI became the first officially confirmed downstream victim of TeamPCP's campaign, stating it was breached as a direct result of the March 24 LiteLLM compromise. This marked the first public victim disclosure linking real-world downstream impact to the supply-chain intrusion.
By 2026-03-31, reporting said LAPSUS$ had released alleged AstraZeneca data for free after failing to sell it, and Cybernews partially verified the dump. AstraZeneca had not publicly commented at the time, marking an escalation beyond the group's earlier breach claim.
On 2026-03-31, Appian published a knowledge-base article addressing the TeamPCP / CanisterWorm supply-chain compromise. This represents an official vendor response and indicates Appian was assessing or disclosing potential impact from the broader campaign.
On 2026-03-31, ownCloud disclosed that its build infrastructure was affected by CVE-2026-33634 stemming from the Trivy supply-chain compromise. The company said customer data and source code were not impacted.
On 2026-03-31, Databricks said via its verified security account that it investigated allegations that TeamPCP-stolen credentials were used against its environment but found nothing in its internal systems. The company requested additional information, marking a public response to a potential downstream impact claim.
On 2026-03-31, reporting said credentials stolen through the Trivy supply-chain compromise were used to access Cisco’s internal development environment via a malicious GitHub Action plugin. This added Cisco as a newly identified downstream victim in the broader TeamPCP-linked campaign.
By 2026-03-31, reporting indicated TeamPCP was operating its own CipherForce ransomware channel in addition to its partnership with the Vect ransomware ecosystem. This added a new attribution detail showing the group pursuing dual ransomware monetization paths rather than relying solely on affiliates.
By March 30, 2026, reporting characterized TeamPCP's activity as a coordinated campaign affecting Trivy, Checkmarx KICS, LiteLLM, and Telnyx across GitHub Actions, Docker Hub, PyPI, npm, and OpenVSX. The campaign was described as using stolen developer and CI/CD credentials to propagate across trusted software ecosystems.
On 2026-03-28, ShinyHunters reportedly published data stolen from the European Commission’s Europa AWS hosting platform after the Trivy-linked compromise. CERT-EU said the breach involved 340 GB of exfiltrated data, including about 52,000 email-related files affecting 71 clients.
On 2026-03-28, new defensive and technical reporting on TeamPCP emerged, including Palo Alto Networks behavioral detection rules for CI/CD attack patterns and additional analysis of the campaign’s Kubernetes wiper and credential fan-out. The update also noted no newly confirmed package compromises in the prior 48 hours, suggesting a temporary pause in expansion while monetization activity continued.
On 2026-03-27, LAPSUS$ publicly claimed a 3GB breach of AstraZeneca allegedly obtained using credentials linked to TeamPCP's supply-chain campaign. AstraZeneca had not confirmed the claim at the time of publication.
On March 27, 2026, attackers used valid credentials to publish malicious Telnyx releases to PyPI. The packages reportedly executed code on import and hid a second-stage payload inside a WAV file.
On 2026-03-27, updated reporting said the March 24 LiteLLM compromise began through the personal GitHub account of CEO Krish Dholakia. This added a new attribution detail about how attackers obtained access in the LiteLLM portion of the TeamPCP campaign.
Posts attributed to the Vect ransomware group on BreachForums claimed a partnership with TeamPCP to turn recent supply-chain compromises into ransomware access operations. Vect was described as an emerging ransomware-as-a-service group seeking affiliates and offering affiliation keys and support.
Reporting cited in the references says TeamPCP started recruiting negotiators after the Trivy compromise, indicating a shift from initial access and supply-chain intrusion toward monetization. This was presented as an early sign that the campaign could evolve into ransomware operations.
On 2026-03-25, Vect Ransomware and TeamPCP allegedly claimed on a dark web forum that they had breached Sportradar AG via the Trivy supply-chain vector and exfiltrated corporate and client data. The post offered the data for sale for up to $50,000 and said it included PII, business records, and third-party credentials linked to FIBA and Bet365.
On 2026-03-25, a public GitHub repository documented TeamPCP's supply-chain campaign with concrete indicators of compromise, including affected versions, hashes, domains, IPs, persistence paths, and malware behavior across GitHub Actions, Docker Hub, OpenVSX, PyPI, and npm. The write-up also provided Microsoft Defender XDR detection guidance and described the npm malware cluster dubbed CanisterWorm, including its use of Internet Computer Protocol canisters for command-and-control.
On March 24, 2026, TeamPCP compromised LiteLLM as part of its supply-chain campaign. Later reporting linked this intrusion to the theft of large numbers of credentials, including the token believed to have enabled the subsequent Telnyx PyPI compromise.
On 2026-03-23, attackers retagged 35 Checkmarx KICS GitHub Action releases to malicious commits, causing users pulling affected tags between 12:58 and 16:50 UTC to receive credential-stealing malware. The same reporting said compromised Checkmarx OpenVSX extensions also delivered second-stage malware and persistence, and attributed the activity to TeamPCP with high confidence based on shared tactics and the same RSA key seen in the earlier Trivy incident.
On 2026-03-22, attackers published malicious Trivy container images 0.69.4, 0.69.5, and 0.69.6 to Docker Hub after compromising Aqua Security's GitHub environment, including force-pushed tags and repository defacements. Aqua Security revoked compromised credentials and removed the malicious images and affected GitHub content, establishing the core Trivy supply-chain compromise event.
On 2026-03-19, attackers reportedly used the compromised Trivy scanner on the European Commission's Europa AWS hosting platform to steal AWS API keys. Later CERT-EU reporting said this intrusion led to large-scale data theft affecting 71 clients, establishing the initial breach event behind the Commission disclosure.
On 2026-03-19, Aqua Security said attackers used compromised credentials to renew the Trivy supply-chain compromise, publishing a malicious Trivy v0.69.4 release and retagging aquasecurity/trivy-action and aquasecurity/setup-trivy to credential-stealing malware. Aqua described it as a continuation of the late-February attack enabled by incomplete post-disclosure credential rotation, and later identified safe versions and rotation guidance for affected users.
On 2026-02-27, a Pwn Request attack reportedly hit Trivy infrastructure, and incomplete remediation allegedly left conditions that enabled TeamPCP's later March 2026 supply-chain campaign. This establishes an earlier precursor event behind the subsequent Trivy-linked compromises.
On 2025-12-31, the Vect ransomware-as-a-service operation reportedly emerged on a Russian-language cybercrime forum. The group was described as rapidly expanding its affiliate base through low-cost access, CIS fee waivers, and a later BreachForums partnership.
In late 2025, TeamPCP was reportedly exploiting exposed Docker and Kubernetes environments as part of its earlier cloud-native criminal activity. This predates the group's March 2026 software supply-chain campaign and shows its operations began with direct cloud and container intrusions before expanding into ecosystem-wide compromises.
In September 2025, TeamPCP reportedly began with relatively unsophisticated attacks against misconfigured Kubernetes clusters. The activity was described as the group's early cloud-native intrusion phase before its later software supply-chain campaign.
Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.
50 references tracked. Mallory keeps watching after this page renders.
itpro.com
Open sourcecysecurity.news
Open sourcecyberscoop.com
Open sourcemalware.news
Open sourcemalware.news
Open sourceironcastle.net
Open sourcethesmallbusinesscybersecurityguy.co.uk
Open sourceisc.sans.edu
Open sourceMap indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.